RPS Report FINAL for web_drb w nancy.indd

Transcription

Appendix A:
Supplement to the Economic Assessment
Edward J. Bloustein School of Planning and Public Policy
Rutgers, The State University of New Jersey
77
SUPPLEMENT TO THE ECONOMIC ASSESSMENT
There are many ways that increasing the size of New Jersey’s renewable energy
portfolio can affect the environment and the economy. One partial equilibrium outcome
is that the unadjusted price of electrical energy will increase. This is because
conventional sources have been selected by the market economy since they are perceived
to be less expensive than alternative fuels, at least given the current electricity generation
set in place. On the other hand, a point of enhancing the renewable portfolio is to meet
environmental objectives. Thus, the price increase of electricity is a prime component of
the costs for meeting those of objectives. Such impacts are discussed elsewhere in detail
using an economic forecast as the modeling vehicle. But there are other potential costs
and benefits of converting to electricity generation based on renewable energy sources.
Indeed, while the relative costs of expanding investment in a renewable portfolio is
embedded in average electricity prices, the relative benefits of the investments are not
counted.
The benefits of a specific investment are typically measured in terms of the jobs
and wealth that are created in the wake of that investment. Increases in these measures
occur due to the augmentation of generation technology production, the installation of the
plant, and the operation and maintenance of the facility. Naturally, to identify the changes
of these measures due to conversion to a new technology, the benefits that would accrue
to the new technology must be compared to the set of benefits that would have accrued
had no policy change taken place. We assumed here, as throughout this report, that
natural gas would serve as the prime conventional energy resource and that gas turbines
would be used to generate electricity. Moreover, we presumed that conventional sources
constructed in the future would be located out of state.
The net investment benefits of two renewable fuel technologies—landfill biogas
and biomass—were acknowledged as negligible. Electricity produced from landfill
biogas was perceived at best to maintain its current low-level production and perhaps
even decline through the end of the study period. Gasified and fluidized-bed biomass
technologies were recognized to be insufficiently price-competitive by 2020 (Navigant,
2004, p. 235): hence, they are supposed to make insignificant into New Jersey’s energy
supplies through 2020. While direct biomass conversion is presently limited, the
technology being used is, in any case, largely conventional. Moreover, while policies
could be developed to induce the location of such facilities in New Jersey, those same
policies could serve to subsidize their eventual conversion to conventional power
facilities. Hence, the economic benefits of both biomass technologies were not
investigated.
The two remaining renewable energy sources for which economic benefits are
considered are (1) solar photovoltaics and (2) wind. The remainder of this section
discusses how their economic benefits were estimated.
In order to estimate the benefits to the New Jersey economy of these two
technologies, the engineering costs of the manufacture and installation of the two
78
Economic Impact Analysis of New Jersey’s Proposed 20% Renewable Portfolio Standard
Center for Energy, Economic & Environmental Policy
technologies had to be identified. Those for solar photovoltaics are displayed in Table
A.1. It was presumed that the cost structure for photovoltaics systems connected with
commercial, industrial, and civic/institutional buildings would not be much different.
Engineering cost estimates for a wind park are shown in Table A.2.
Table A.1: Cost Breakdown of a Typical 8 KW
Residential Solar Photovoltaic Array
Material/Service
Cost in 2004 Dollars
Marketing and Sales
3,200
Engineering and Design
480
Post-installation Servicing
960
General and Administrative
2,400
Modules
29,280
Inverters
7,680
Mounting
3,360
Miscellaneous Material
1,920
Array Labor
5,360
Electrician Labor
3,680
Total
58,320
Source: Lyle Rawlings, Advanced Solar Products, Hopewell, New Jersey.
Table A.2: An Example Cost Analysis for a 60 MW Wind Park
Item
Cost in 2004 Dollars
40 1.5 MW turbines
46,640,000
Site preparation & grid connection
9,148,000
Interest and contingencies
Project development & feasibility study
3,514,000
965,000
Engineering
611,000
Total
60,878,000
Source: Masters (2004, Table 6.8, p. 373).
With only minor modification to align the materials, services and other items so
they were aligned with model sectors, the data in Tables A.1 and A.2 were entered into
the R/Econ I-O model for the State of New Jersey.
For each of the two technologies, two different R/Econ I-O model runs were
undertaken. In the case of PV, we first assumed that the technology was only installed in
New Jersey. Prior experience has revealed that on average 90 percent of all contracting
79
work in New Jersey is performed by contractors who live in the state. Hence, we applied
this proportion in the present study. Table A.3 shows a summary of the results on the
economic and tax impacts on New Jersey of 8 Mw of PV installation in the State. Note
that the direct output effects (the first column of II.1 in Table A.3) are not 1,000 times the
total spending in Table A.1. This is because the PV units are assumed to be produced in
New Jersey only in so much as they would normally meet state-based PV demand.
Needless to say, a substantial portion of spending on solar photovoltaic units currently
tends to take place outside of the State. Indeed, based on the results shown in Table A.3,
only about a third ($19 million) of the $58 million initial investment is assumed to be
spent on goods and services within the State under current circumstances, and more than
half of this is for construction services required to install the arrays in place within the
State. Moreover, looking at the bottom of Table A.3, it is clear that $1 million of 2002
dollars spent on installed PV will yield 3.8 full-time-equivalent jobs, about $280,000 in
total state wealth (gross state product), $210,000 in earnings to state workers, and a total
of $19,000 split between state and local tax revenues. Generally speaking, these are lessthan-stellar returns to a state-based infrastructure investment. On the other hand, the
average annual earnings per job are estimated to be $54,700, 15% above the average for
the state ($47,420 in 2002). Of course, environmental savings of PV compared to
conventional fuels are not included in these calculations.
The second run for PV assumes a policy that invokes a mandate that all PV
installed in New Jersey is also produced in the State. Hence, not only were 90 percent of
contractors New Jerseyans, but all design, marketing, sales, and production aspects of the
PV installed in New Jersey buildings also are presumed to take place in the State. Table
A.4 summarizes the economic impacts on the State of bring this policy into play. In
general, the effects of this New Jersey-only technology policy yields an economic impact
that is nearly two and a half times larger than if no such policy was put in place. In this
case, each million spent on PV installations will produce about 8.7 jobs, $700,000 in
wealth, $516,000 in earnings, and $55,000 in state and local taxes. While these impacts
are low in terms of jobs, they comport well otherwise with other state-based
infrastructure investments. Moreover, the earnings per jobs are estimated to be even
higher than in the base case—at $59,600, 10% higher…or 25% above the state’s annual
average in 2002.
In the case of wind power, the basic assumption was that wind power demand
would be met from sources outside of New Jersey but within the PJM market. Hence,
increasing wind’s share of the RPS was understood to have negligible economic impacts
upon the state without prompting the industry with incentives. The other alternative is to
apply policies that not only induce the installation of wind-based electric power plants
within the State in the form of offshore turbines, but also to induce the production of the
turbines and towers themselves. The economic impacts of this recourse are displayed in
Table A.5. In general, the economic impacts per $1 million of installing and producing
wind power are just slightly lower than those for PV produced in the State: 7.6 jobs,
$636,000 in wealth, $462,000 in earnings, and $53,000 in state and local taxes. Due to its
larger yield in manufacturing jobs, it impacts include jobs with average annual earnings
of $60,600, slightly above those for State-based PV.
80
Table A.3: Economic and Tax Impacts on New Jersey
of Installing 8 MW of Residential Photovoltaic Systems
(Year 2000 Dollars)
Output
(000$)
Economic Component
Employment
(jobs)
Income
(000$)
Gross State
Product (000$)
I. TOTAL EFFECTS (Direct and Indirect/Induced)*
1. Agriculture
2. Agri. Serv., Forestry, & Fish
3. Mining
4. Construction
5. Manufacturing
6. Transport. & Public Utilities
7. Wholesale
8. Retail Trade
9. Finance, Ins., & Real Estate
10. Services
Private Subtotal
Public
11. Government
Total Effects (Private and Public)
24.9
11.6
6.5
10,243.8
5,553.9
1,034.3
693.2
1,550.1
1,664.8
6,726.6
27,509.7
0
0
0
106
25
4
4
25
9
49
222
2.6
5.7
2.3
6,504.4
1,467.9
263.8
281.9
581.2
554.6
2,451.4
12,115.9
4.9
9.6
4.3
8,482.6
1,776.0
415.0
297.7
907.0
1,139.8
3,338.4
16,375.3
58.5
27,568.3
0
222
17.9
12,133.8
28.5
16,403.7
19,113.5
8,454.8
27,568.3
1.442
156
66
222
1.427
9,326.1
2,807.6
12,133.8
1.301
12,246.5
4,157.3
16,403.7
1.339
II. DISTRIBUTION OF EFFECTS/MULTIPLIER
1. Direct Effects
2. Indirect and Induced Effects
3. Total Effects
4. Multipliers (3/1)
III. COMPOSITION OF GROSS STATE PRODUCT
11,230.0
2,204.7
276.3
252.9
1,675.5
374.6
1,300.9
2,969.0
16,403.7
1. Wages--Net of Taxes
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
3. Profits, dividends, rents, and other
4. Total Gross State Product (1+2+3)
IV. TAX ACCOUNTS
1. Income --Net of Taxes
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
Business
Household
Total
11,230.0
2,204.7
276.3
252.9
1,675.5
374.6
1,300.9
0.0
2,462.1
315.6
276.3
1,870.2
1,870.2
0.0
4,666.8
591.9
529.2
3,545.7
2,244.8
1,300.9
EFFECTS PER MILLION DOLLARS OF INITIAL EXPENDITURE
3.8
208,055.4
9,073.4
10,149.5
281,270.8
Employment (Jobs)
Income
State Taxes
Local Taxes
Gross State Product
INITIAL EXPENDITURE IN DOLLARS
58,320,000.0
Note: Detail may not sum to totals due to rounding.
*Terms:
Direct Effects --the proportion of direct spending on goods and services produced in the specified region.
Indirect Effects--the value of goods and services needed to support the provision of those direct economic effects.
Induced Effects--the value of goods and services needed by households that provide the direct and indirect labor.
81
of Installing 8 MW of Residential Photovoltaic Systems,
Assuming All Production Takes Place in the State
(Year 2000 Dollars)
Output
(000$)
Economic Component
Employment
(jobs)
Income
(000$)
Gross State
Product (000$)
1. Agriculture
3. Mining
4. Construction
5. Manufacturing
7. Wholesale
8. Retail Trade
10. Services
Private Subtotal
Public
11. Government
60.4
47.9
9.4
11,173.3
46,388.8
2,960.6
2,645.1
3,767.7
4,375.3
11,454.2
82,882.7
0
1
0
108
193
11
15
61
22
93
504
6.4
24.7
3.2
6,632.8
14,318.2
724.8
1,075.6
1,408.6
1,427.9
4,432.3
30,054.6
11.7
40.2
6.1
8,792.5
18,542.0
1,180.8
1,136.1
2,192.6
3,005.2
5,691.5
40,598.7
180.1
83,062.9
1
505
55.3
30,109.9
89.3
40,687.9
58,320.0
24,742.9
83,062.9
1.424
319
186
505
1.582
21,894.6
8,215.3
30,109.9
1.375
28,733.6
11,954.3
40,687.9
1.416
1. Direct Effects
3. Total Effects
27,650.1
5,688.1
943.7
787.3
3,957.1
728.9
3,228.3
7,349.8
40,687.9
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
IV. TAX ACCOUNTS
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
Business
Household
Total
27,650.1
5,688.1
943.7
787.3
3,957.1
728.9
3,228.3
0.0
6,109.6
783.1
685.6
4,640.8
4,640.8
0.0
11,797.6
1,726.8
1,472.9
8,598.0
5,369.7
3,228.3
Employment (Jobs)
Income
State Taxes
Local Taxes
Gross State Product
*Terms:
82
8.7
516,287.6
25,254.9
29,609.3
697,666.5
58,320,000.0
of Installing 60 MW Off-shore Wind Park,
Assuming All Production Takes Place in the State
(Year 2000 Dollars)
Output
(000$)
Economic Component
Employment
(jobs)
Income
(000$)
Gross State
Product (000$)
1. Agriculture
3. Mining
4. Construction
5. Manufacturing
7. Wholesale
8. Retail Trade
10. Services
Private Subtotal
Public
11. Government
60.2
46.7
108.3
6,601.3
56,821.3
7,360.5
2,999.4
3,704.1
4,240.8
6,510.1
88,452.7
0
1
1
40
240
21
17
60
22
62
463
6.2
24.1
37.3
3,066.3
16,624.7
1,457.5
1,219.7
1,382.6
1,420.2
2,838.8
28,077.5
11.5
39.2
71.1
4,340.5
21,836.4
2,812.5
1,288.3
2,150.1
2,854.4
3,233.3
38,637.1
218.5
88,671.2
1
464
66.6
28,144.0
105.3
38,742.4
60,878.0
27,793.2
88,671.2
1.457
262
203
464
1.775
18,727.9
9,416.1
28,144.0
1.503
25,558.5
13,183.9
38,742.4
1.516
1. Direct Effects
3. Total Effects
25,637.1
5,524.4
1,033.9
841.2
3,649.3
631.8
3,017.5
7,580.9
38,742.4
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
IV. TAX ACCOUNTS
2. Taxes
a. Local
b. State
c. Federal
General
Social Security
Business
Household
Total
25,637.1
5,524.4
1,033.9
841.2
3,649.3
631.8
3,017.5
0.0
5,710.7
732.0
640.8
4,337.8
4,337.8
0.0
11,235.1
1,765.9
1,482.1
7,987.1
4,969.6
3,017.5
7.6
462,302.2
24,344.8
29,007.5
636,394.0
Employment (Jobs)
Income
State Taxes
Local Taxes
Gross State Product
60,878,000
*Terms:
83
84
$0.100
$0.111
$0.095
73,644
26,538
47,106
$7.4
$2.9
$4.5
$932.6
$822.6
$630.1
$192.6
$110.0
$0.100
$0.111
$0.093
66,537
26,367
40,170
$6.7
$2.9
$3.7
$1,108.3
$858.8
$697.1
$161.7
$249.5
Electric Price per Kilowatt Hour
Electricity Usage in 1000 Megawatt Hours
$8.16
$364.10
15.3
8.8
6.6
4079.7
199.4
3.40
0.90
2.50
$8.04
15.4
8.9
6.4
4030.2
196.6
3.25
0.75
2.50
NonAgricultural Employment (Thousands)
NJ Consumer Price Index (1982-84=100) 1
Renewable Portfolio Standard
4.13
1.63
2.50
201.8
4133.8
15.4
8.7
6.7
$373.83
$8.40
$897.7
$842.7
$639.4
$203.3
$55.0
$7.6
$3.0
$4.6
75,482
26,946
48,536
$0.100
$0.110
$0.095
2006
1 The NJ CPI is a population-weighted average of the CPIs for NY-NJ-CT and PA-NJ.
Class 2:Hydroelectric and Waste-to-Energy
Class 1: Photovoltaics, et al.
Other
Electric Utitliies
Employment at Utilities (Thousands)
Gross State Product ($Billions 2000=100)
Gross State Product for Utilities ($ Billions)
$354.57
Transitional Facility Assessment
Corporate Business
Sales
Energy Taxes - TEFA
Energy Taxes ($ Millions)
Other
Residential
Electric Revenues ($ Billions)
Other
Residential
Other
Residential
2005
2004
Electric Utilities: Prices, Usage, Taxes, etc.
5.12
2.62
2.50
205.3
4183.6
15.4
8.6
6.8
$383.97
$8.52
$856.7
$856.7
$647.8
$208.9
$0.0
$7.7
$3.0
$4.7
76,615
27,405
49,210
$0.100
$0.110
$0.095
2007
6.21
3.71
2.50
209.5
4227.1
15.4
8.6
6.8
$393.62
$8.56
$869.4
$869.4
$658.5
$210.9
$0.0
$7.9
$3.1
$4.8
77,620
27,864
49,756
$0.101
$0.110
$0.097
2008
6.50
4.00
2.50
213.9
4270.3
15.3
8.5
6.8
$403.29
$8.63
$881.4
$881.4
$667.4
$214.0
$0.0
$8.0
$3.1
$4.9
78,629
28,315
50,314
$0.102
$0.111
$0.097
2009
6.50
4.00
2.50
218.7
4317.4
15.3
8.5
6.9
$414.16
$8.72
$893.8
$893.8
$675.7
$218.1
$0.0
$8.2
$3.2
$5.0
79,635
28,758
50,877
$0.102
$0.111
$0.098
2010
6.50
4.00
2.50
224.2
4363.2
15.3
8.4
6.9
$426.09
$8.82
$907.2
$907.2
$684.4
$222.9
$0.0
$8.3
$3.3
$5.0
80,621
29,195
51,425
$0.103
$0.112
$0.098
2011
6.50
4.00
2.50
229.7
4411.9
15.3
8.4
6.9
$438.96
$8.95
$922.2
$922.2
$693.6
$228.6
$0.0
$8.5
$3.3
$5.1
81,636
29,632
52,004
$0.103
$0.112
$0.099
2012
6.50
4.00
2.50
235.3
4463.4
15.3
8.4
6.9
$453.03
$9.09
$938.4
$938.4
$703.0
$235.4
$0.0
$8.6
$3.4
$5.2
82,679
30,073
52,607
$0.104
$0.113
$0.099
2013
6.50
4.00
2.50
241.1
4521.9
15.2
8.4
6.9
$468.20
$9.25
$955.6
$955.6
$713.2
$242.5
$0.0
$8.8
$3.4
$5.3
83,794
30,522
53,271
$0.105
$0.113
$0.100
2014
6.50
4.00
2.50
247.1
4576.9
15.2
8.3
6.9
$484.17
$9.41
$973.5
$973.5
$723.7
$249.8
$0.0
$8.9
$3.5
$5.4
84,952
30,982
53,970
$0.105
$0.114
$0.100
2015
Table A.6: R/ECON BPU BASELINE FORECAST
6.50
4.00
2.50
253.2
4628.1
15.2
8.3
6.9
$500.86
$9.57
$991.3
$991.3
$734.1
$257.2
$0.0
$9.1
$3.6
$5.5
86,064
31,450
54,614
$0.106
$0.115
$0.101
2016
6.50
4.00
2.50
259.4
4677.9
15.2
8.3
6.9
$518.30
$9.72
$1,008.5
$1,008.5
$744.5
$264.0
$0.0
$9.3
$3.7
$5.6
87,151
31,923
55,228
$0.106
$0.115
$0.102
2017
6.50
4.00
2.50
265.8
4730.5
15.2
8.3
6.9
$536.48
$9.86
$1,025.4
$1,025.4
$755.1
$270.3
$0.0
$9.5
$3.8
$5.7
88,242
32,397
55,845
$0.107
$0.116
$0.102
2018
6.50
4.00
2.50
272.4
4784.8
15.2
8.3
6.9
$555.63
$9.99
$1,042.5
$1,042.5
$766.1
$276.4
$0.0
$9.7
$3.8
$5.8
89,368
32,872
56,496
$0.108
$0.117
$0.103
2019
6.50
4.00
2.50
279.1
4845.0
15.2
8.3
6.9
$575.88
$10.12
$1,060.3
$1,060.3
$777.6
$282.7
$0.0
$9.8
$3.9
$5.9
90,539
33,350
57,189
$0.108
$0.117
$0.104
2020
100.0%
433.3%
0.0%
41.9%
20.2%
-1.2%
-6.4%
7.7%
62.4%
26.0%
-4.3%
23.5%
11.5%
74.8%
47.8%
33.3%
59.2%
36.1%
26.5%
42.4%
8.0%
5.4%
11.8%
Growth
2004 - 2020
This part of Appendix A includes the BASELINE forecast used in this study as well as 4 other
scenarios discussed in Chapter 2.
85
199.4
3.40
0.90
2.50
196.8
3.25
0.75
2.50
4.13
1.63
2.50
201.8
4133.7
15.4
8.7
6.7
$373.80
$8.39
$897.6
$842.6
$639.4
$203.2
$55.0
$7.6
$3.0
$4.6
75,480
26,946
48,535
$0.100
$0.110
$0.095
2006
NJ Consumer Price Index (1982-84=100)
1
Other
Electric Utitliies
4079.6
$8.16
$8.04
Corporate Business
4030.1
$932.6
$822.6
$630.1
$192.5
$110.0
$1,078.8
$858.8
$697.1
$161.7
$220.0
Sales
Total - TEFA
Other
15.3
8.8
6.6
$7.4
$2.9
$4.5
$6.7
$2.9
$3.7
Residential
Other
Residential
15.4
8.9
6.4
73,643
26,538
47,105
66,536
26,367
40,169
Other
Residential
$364.07
$0.100
$0.111
$0.095
$0.100
$0.111
$0.093
$354.56
2005
2004
5.12
2.62
2.50
205.3
4183.5
15.4
8.6
6.8
$383.94
$8.52
$856.6
$856.6
$647.8
$208.8
$0.0
$7.7
$3.0
$4.7
76,614
27,405
49,209
$0.100
$0.110
$0.095
2007
6.21
3.71
2.50
209.5
4227.0
15.4
8.6
6.8
$393.60
$8.56
$869.5
$869.5
$658.5
$211.0
$0.0
$7.9
$3.1
$4.8
77,619
27,863
49,755
$0.101
$0.110
$0.097
2008
7.22
4.72
2.50
214.0
4270.2
15.3
8.5
6.8
$403.28
$8.63
$884.8
$884.8
$670.8
$214.0
$0.0
$8.1
$3.1
$4.9
78,552
28,308
50,244
$0.103
$0.111
$0.098
2009
8.44
5.94
2.50
218.9
4317.3
15.3
8.5
6.9
$414.13
$8.72
$904.6
$904.6
$686.5
$218.1
$0.0
$8.3
$3.2
$5.1
79,516
28,744
50,772
$0.105
$0.112
$0.100
2010
9.67
7.17
2.50
224.5
4363.0
15.3
8.4
6.9
$426.02
$8.82
$927.3
$927.3
$704.6
$222.7
$0.0
$8.6
$3.3
$5.3
80,484
29,177
51,307
$0.107
$0.114
$0.103
2011
10.90
8.40
2.50
230.2
4411.5
15.3
8.4
6.9
$438.82
$8.94
$952.9
$952.9
$724.5
$228.4
$0.0
$9.0
$3.5
$5.5
81,489
29,610
51,879
$0.110
$0.117
$0.106
2012
12.13
9.63
2.50
235.9
4462.8
15.3
8.4
6.9
$452.81
$9.09
$980.8
$980.8
$745.8
$235.0
$0.0
$9.3
$3.6
$5.7
82,527
30,049
52,479
$0.113
$0.119
$0.109
2013
13.35
10.85
2.50
241.7
4521.2
15.2
8.4
6.9
$467.91
$9.24
$1,010.6
$1,010.6
$768.5
$242.1
$0.0
$9.7
$3.7
$6.0
83,639
30,497
53,142
$0.116
$0.122
$0.112
2014
14.58
12.08
2.50
247.8
4576.0
15.2
8.3
6.9
$483.79
$9.40
$1,041.8
$1,041.8
$792.4
$249.4
$0.0
$10.1
$3.9
$6.2
84,797
30,956
53,841
$0.119
$0.125
$0.115
2015
Table A.7: R/ECON BPU STRAIGHTLINE RPS INCREASE
2009 to 20% in 2020
15.81
13.31
2.50
254.0
4627.0
15.2
8.3
6.9
$500.38
$9.56
$1,073.4
$1,073.4
$816.6
$256.8
$0.0
$10.5
$4.0
$6.5
85,909
31,423
54,486
$0.122
$0.127
$0.119
2016
17.03
14.53
2.50
260.3
4676.6
15.2
8.3
6.9
$517.73
$9.71
$1,105.0
$1,105.0
$841.3
$263.6
$0.0
$10.9
$4.1
$6.7
86,996
31,895
55,101
$0.125
$0.130
$0.122
2017
18.26
15.76
2.50
266.8
4729.0
15.2
8.3
6.9
$535.83
$9.85
$1,136.7
$1,136.7
$866.7
$270.0
$0.0
$11.3
$4.3
$7.0
88,089
32,369
55,720
$0.128
$0.133
$0.125
2018
19.49
16.99
2.50
273.5
4783.0
15.2
8.3
6.9
$554.89
$9.98
$1,169.0
$1,169.0
$892.9
$276.1
$0.0
$11.7
$4.5
$7.3
89,216
32,844
56,371
$0.131
$0.136
$0.129
2019
20.00
17.50
2.50
280.2
4843.0
15.2
8.3
6.9
$575.05
$10.12
$1,198.5
$1,198.5
$916.1
$282.4
$0.0
$12.1
$4.6
$7.5
90,444
33,327
57,117
$0.134
$0.138
$0.131
2020
86
$8.16
$364.09
15.3
8.8
6.6
4079.7
$8.04
$354.57
15.4
8.9
6.4
4030.2
4.13
1.63
2.50
201.9
4133.8
15.4
8.7
6.7
$373.80
$8.37
$899.7
$844.7
$642.6
$202.1
$55.0
$7.6
$3.0
$4.7
75,441
26,940
48,501
$0.101
$0.111
$0.096
2006
1
Other
Electric Utitliies
Corporate Business
Sales
3.40
0.90
2.50
$933.0
$823.0
$630.6
$192.4
$110.0
$1,108.3
$858.8
$697.1
$161.7
$249.5
Energy Taxes - TEFA
3.25
0.75
2.50
$7.4
$2.9
$4.5
$6.7
$2.9
$3.7
Other
Residential
Other
Residential
Other
199.4
73,628
26,537
47,091
66,537
26,367
40,170
196.6
$0.101
$0.111
$0.095
$0.100
$0.111
$0.093
Residential
2005
2004
5.12
2.62
2.50
205.4
4183.6
15.4
8.6
6.8
$383.92
$8.47
$860.7
$860.7
$653.9
$206.8
$0.0
$7.8
$3.0
$4.8
76,576
27,398
49,178
$0.102
$0.111
$0.097
2007
6.21
3.71
2.50
209.5
4227.1
15.4
8.6
6.8
$393.57
$8.52
$876.6
$876.6
$667.5
$209.1
$0.0
$8.0
$3.1
$4.9
77,582
27,856
49,726
$0.103
$0.112
$0.099
2008
6.50
4.00
2.50
213.9
4270.2
15.3
8.5
6.8
$403.25
$8.60
$891.7
$891.7
$679.0
$212.7
$0.0
$8.2
$3.2
$5.0
78,600
28,309
50,291
$0.104
$0.113
$0.100
2009
6.50
4.00
2.50
218.7
4317.4
15.3
8.5
6.9
$414.14
$8.70
$906.7
$906.7
$689.3
$217.4
$0.0
$8.4
$3.3
$5.1
79,613
28,753
50,860
$0.105
$0.114
$0.100
2010
6.50
4.00
2.50
224.2
4363.2
15.3
8.4
6.9
$426.08
$8.82
$922.1
$922.1
$699.5
$222.6
$0.0
$8.6
$3.4
$5.2
80,608
29,192
51,416
$0.106
$0.115
$0.101
2011
6.50
4.00
2.50
229.8
4411.9
15.3
8.4
6.9
$438.96
$8.95
$938.5
$938.5
$709.9
$228.6
$0.0
$8.7
$3.4
$5.3
81,625
29,630
51,995
$0.107
$0.116
$0.102
2012
6.50
4.00
2.50
235.3
4463.4
15.3
8.4
6.9
$453.02
$9.09
$955.9
$955.9
$720.5
$235.3
$0.0
$8.9
$3.5
$5.4
82,671
30,071
52,600
$0.107
$0.116
$0.103
2013
6.50
4.00
2.50
241.1
4521.9
15.2
8.4
6.9
$468.20
$9.25
$974.2
$974.2
$731.8
$242.4
$0.0
$9.1
$3.6
$5.5
83,786
30,521
53,265
$0.108
$0.117
$0.103
2014
6.50
4.00
2.50
247.1
4576.9
15.2
8.3
6.9
$484.16
$9.41
$993.3
$993.3
$743.6
$249.7
$0.0
$9.3
$3.7
$5.6
84,943
30,981
53,963
$0.109
$0.118
$0.104
2015
Table A.8: R/ECON BPU BASELINE WITH HIGH PPI ENERGY
6.50
4.00
2.50
253.2
4628.1
15.2
8.3
6.9
$500.85
$9.56
$1,012.3
$1,012.3
$755.3
$257.0
$0.0
$9.5
$3.7
$5.7
86,055
31,449
54,606
$0.110
$0.119
$0.105
2016
6.50
4.00
2.50
259.4
4677.9
15.2
8.3
6.9
$518.29
$9.71
$1,030.9
$1,030.9
$767.1
$263.7
$0.0
$9.7
$3.8
$5.9
87,141
31,921
55,220
$0.111
$0.120
$0.106
2017
6.50
4.00
2.50
265.8
4730.5
15.2
8.3
6.9
$536.47
$9.85
$1,049.3
$1,049.3
$779.2
$270.1
$0.0
$9.9
$3.9
$6.0
88,233
32,396
55,837
$0.112
$0.121
$0.107
2018
6.50
4.00
2.50
272.4
4784.8
15.2
8.3
6.9
$555.62
$9.98
$1,068.0
$1,068.0
$791.8
$276.2
$0.0
$10.1
$4.0
$6.1
89,358
32,871
56,487
$0.112
$0.122
$0.108
2019
6.50
4.00
2.50
279.1
4845.0
15.2
8.3
6.9
$575.88
$10.12
$1,087.4
$1,087.4
$805.0
$282.5
$0.0
$10.3
$4.1
$6.2
90,529
33,348
57,181
$0.113
$0.123
$0.109
2020
100.0%
433.3%
0.0%
41.9%
20.2%
-1.2%
-6.4%
7.7%
62.4%
25.9%
-1.9%
26.6%
15.5%
74.6%
55.1%
40.2%
66.8%
36.1%
26.5%
42.3%
13.0%
10.8%
17.2%
Growth
2004 - 2020
87
199.4
3.40
0.90
2.50
3.25
0.75
2.50
4079.8
196.6
4030.2
15.3
8.8
6.6
1
Other
Electric Utitliies
Gross State Product for Utilities ($ Billions 2000=100)
Corporate Business
Sales
Total - TEFA
15.4
8.9
6.4
$932.6
$822.6
$630.1
$192.6
$110.0
$1,108.3
$858.8
$697.1
$161.7
$249.5
Other
$8.26
$7.4
$2.9
$4.5
$6.7
$2.9
$3.7
Residential
Other
Residential
$391.04
73,645
26,538
47,107
66,537
26,367
40,170
Other
Residential
$8.13
$0.100
$0.111
$0.095
$0.100
$0.111
$0.093
$380.81
2005
2004
4.13
1.63
2.50
201.8
4134.0
15.4
8.7
6.7
$401.50
$8.50
$897.7
$842.7
$639.4
$203.3
$55.0
$7.6
$3.0
$4.6
75,484
26,946
48,538
$0.100
$0.110
$0.095
2006
5.12
2.62
2.50
205.3
4183.9
15.4
8.6
6.8
$412.39
$8.62
$856.7
$856.7
$647.8
$208.9
$0.0
$7.7
$3.0
$4.7
76,618
27,405
49,213
$0.100
$0.110
$0.095
2007
6.21
3.71
2.50
209.5
4227.6
15.4
8.6
6.8
$422.76
$8.67
$869.5
$869.5
$658.5
$211.0
$0.0
$7.9
$3.1
$4.8
77,625
27,864
49,761
$0.101
$0.110
$0.097
2008
7.22
4.72
2.50
213.9
4271.4
15.3
8.5
6.8
$433.15
$8.73
$881.9
$881.9
$667.9
$214.0
$0.0
$8.0
$3.1
$4.9
78,611
28,308
50,302
$0.102
$0.111
$0.097
2009
8.44
5.94
2.50
218.8
4317.9
15.3
8.5
6.9
$444.83
$8.82
$894.9
$894.9
$676.8
$218.1
$0.0
$8.2
$3.2
$5.0
79,622
28,753
50,869
$0.102
$0.111
$0.098
2010
9.67
7.17
2.50
224.3
4363.5
15.3
8.4
6.9
$457.63
$8.93
$910.9
$910.9
$688.1
$222.9
$0.0
$8.4
$3.3
$5.1
80,566
29,193
51,372
$0.104
$0.112
$0.099
2011
10.90
8.40
2.50
229.9
4412.5
15.3
8.4
6.9
$471.43
$9.05
$927.2
$927.2
$698.7
$228.5
$0.0
$8.5
$3.3
$5.2
81,617
29,625
51,993
$0.104
$0.113
$0.100
2012
12.13
9.63
2.50
235.4
4463.9
15.3
8.4
6.9
$486.52
$9.20
$949.0
$949.0
$713.8
$235.3
$0.0
$8.8
$3.4
$5.4
82,592
30,068
52,524
$0.106
$0.114
$0.102
2013
13.35
10.85
2.50
241.2
4522.5
15.2
8.4
6.9
$502.81
$9.36
$966.8
$966.8
$724.4
$242.4
$0.0
$9.0
$3.5
$5.5
83,781
30,520
53,261
$0.107
$0.114
$0.103
2014
14.58
12.08
2.50
247.2
4577.5
15.2
8.3
6.9
$519.94
$9.52
$986.2
$986.2
$736.4
$249.8
$0.0
$9.1
$3.6
$5.6
84,939
30,974
53,964
$0.108
$0.115
$0.103
2015
Table A.9: R/ECON BPU STRAIGHTLINE RPS INCREASE 2009 to 20% IN 2020
OUT-OF-STATE MANUFACTURING OF RENEWABLE TECHNOLOGIES
15.81
13.31
2.50
253.3
4629.1
15.2
8.3
6.9
$537.86
$9.68
$1,004.5
$1,004.5
$747.4
$257.1
$0.0
$9.3
$3.6
$5.7
86,059
31,444
54,615
$0.108
$0.116
$0.104
2016
17.03
14.53
2.50
259.5
4678.9
15.2
8.3
6.9
$556.60
$9.83
$1,021.0
$1,021.0
$757.0
$264.0
$0.0
$9.5
$3.7
$5.7
87,162
31,920
55,242
$0.109
$0.117
$0.104
2017
18.26
15.76
2.50
265.9
4731.7
15.2
8.3
6.9
$576.13
$9.97
$1,041.0
$1,041.0
$770.7
$270.3
$0.0
$9.7
$3.8
$5.9
88,214
32,390
55,825
$0.110
$0.118
$0.106
2018
19.49
16.99
2.50
272.5
4786.3
15.2
8.3
6.9
$596.71
$10.11
$1,059.8
$1,059.8
$783.3
$276.5
$0.0
$10.0
$3.9
$6.0
89,358
32,867
56,490
$0.111
$0.119
$0.107
2019
20.00
17.50
2.50
279.2
4846.8
15.2
8.3
6.9
$618.47
$10.25
$1,081.3
$1,081.3
$798.6
$282.7
$0.0
$10.2
$4.0
$6.2
90,535
33,347
57,188
$0.112
$0.120
$0.108
2020
88
2005
$0.100
$0.111
$0.095
73,648
26,538
47,110
$7.4
$2.9
$4.5
$932.7
$822.7
$630.1
$192.6
$110.0
$0.100
$0.111
$0.093
66,537
26,367
40,170
$6.7
$2.9
$3.7
$1,108.3
$858.8
$697.1
$161.7
$249.5
$8.26
$391.06
15.3
8.8
6.6
4080.5
199.4
3.40
0.90
2.50
$8.13
15.4
8.9
6.4
4030.2
196.6
3.25
0.75
2.50
NJ Consumer Price Index (1982-84=100) 1
Other
Electric Utitliies
Gross State Product for Utilities ($ Billions 2000=100)
$380.81
Corporate Business
Sales
Total - TEFA
Other
Residential
Other
Residential
Other
Residential
2004
4.13
1.63
2.50
201.8
4134.8
15.4
8.7
6.7
$401.57
$8.50
$897.8
$842.8
$639.5
$203.3
$55.0
$7.6
$3.0
$4.6
75,494
26,946
48,548
$0.100
$0.110
$0.095
2006
5.12
2.62
2.50
205.3
4185.1
15.4
8.6
6.8
$412.50
$8.62
$856.8
$856.8
$647.9
$208.9
$0.0
$7.7
$3.0
$4.7
76,629
27,406
49,223
$0.100
$0.110
$0.095
2007
R/ECON BPU STRAIGHTLINE RPS INCREASE 2009 to 20% IN 2020 WITH MANUFACTURING AND MAINTENANCE OF NEW TECHNOLOGIES INSTATE
6.21
3.71
2.50
209.5
4231.0
15.4
8.6
6.8
$422.93
$8.67
$869.7
$869.7
$658.7
$211.0
$0.0
$7.9
$3.1
$4.8
77,651
27,865
49,786
$0.101
$0.110
$0.097
2008
7.22
4.72
2.50
213.9
4276.4
15.4
8.5
6.8
$433.40
$8.73
$882.3
$882.3
$668.3
$214.1
$0.0
$8.0
$3.1
$4.9
78,664
28,310
50,354
$0.102
$0.111
$0.097
2009
8.44
5.94
2.50
218.8
4321.4
15.3
8.5
6.9
$445.16
$8.83
$895.4
$895.4
$677.2
$218.2
$0.0
$8.2
$3.2
$5.0
79,679
28,756
50,923
$0.102
$0.111
$0.098
2010
9.67
7.17
2.50
224.3
4367.2
15.3
8.4
6.9
$458.02
$8.93
$911.4
$911.4
$688.4
$223.0
$0.0
$8.4
$3.3
$5.1
80,611
29,198
51,413
$0.104
$0.112
$0.099
2011
10.90
8.40
2.50
229.9
4417.3
15.3
8.4
6.9
$471.89
$9.06
$927.7
$927.7
$699.0
$228.7
$0.0
$8.6
$3.3
$5.2
81,674
29,630
52,045
$0.104
$0.113
$0.100
2012
12.13
9.63
2.50
235.4
4469.1
15.3
8.4
6.9
$487.03
$9.20
$949.6
$949.6
$714.2
$235.4
$0.0
$8.8
$3.4
$5.4
82,658
30,073
52,585
$0.106
$0.114
$0.102
2013
13.35
10.85
2.50
241.2
4528.1
15.3
8.4
6.9
$503.37
$9.36
$967.3
$967.3
$724.9
$242.5
$0.0
$9.0
$3.5
$5.5
83,852
30,527
53,326
$0.107
$0.114
$0.103
2014
14.58
12.08
2.50
247.2
4584.2
15.2
8.3
6.9
$520.57
$9.52
$986.8
$986.8
$737.0
$249.8
$0.0
$9.1
$3.6
$5.6
85,019
30,981
54,038
$0.108
$0.115
$0.103
2015
15.81
13.31
2.50
253.3
4636.6
15.2
8.3
6.9
$538.55
$9.69
$1,005.2
$1,005.2
$748.0
$257.2
$0.0
$9.3
$3.6
$5.7
86,153
31,452
54,701
$0.108
$0.116
$0.104
2016
17.03
14.53
2.50
259.5
4686.7
15.2
8.3
6.9
$557.34
$9.84
$1,021.7
$1,021.7
$757.7
$264.1
$0.0
$9.5
$3.7
$5.8
87,263
31,929
55,335
$0.109
$0.117
$0.104
2017
Table A.10: R/ECON BPU STRAIGHTLINE RPS INCREASE 2009 to 20% IN 2020 WITH
MANUFACTURING AND MAINTENANCE OF NEW TECHNOLOGIES INSTATE
18.26
15.76
2.50
265.9
4740.7
15.2
8.3
6.9
$576.95
$9.98
$1,041.9
$1,041.9
$771.4
$270.5
$0.0
$9.8
$3.8
$5.9
88,325
32,399
55,926
$0.110
$0.118
$0.106
2018
19.49
16.99
2.50
272.5
4796.7
15.2
8.3
6.9
$597.61
$10.11
$1,060.7
$1,060.7
$784.2
$276.6
$0.0
$10.0
$3.9
$6.1
89,486
32,878
56,608
$0.111
$0.119
$0.107
2019
20.00
17.50
2.50
279.2
4858.2
15.2
8.3
6.9
$619.4542
$10.25
$1,082.4
$1,082.4
$799.5
$282.9
$0.0
$10.2
$4.0
$6.2
90,679
33,359
57,320
$0.112
$0.120
$0.108
2020
Appendix B:
Supplement to the Environmental
Externality Analysis
89
SUPPLEMENT TO THE ENVIRONMENTAL EXTERNALITY ANALYSIS
1. INTRODUCTION
The purpose of this appendix is to expand upon the environmental externality discussion
in Chapter 3.
This appendix uses a benefit transfer approach. This method consists of conducting a
thorough literature review of recent, North American applicable, peer-reviewed studies.
The key question in following this approach is how closely other studies resemble the
situation in New Jersey based on timeliness, methodology, and other factors. Ideally,
primary research based on the characteristics of New Jersey is preferred to a benefit
transfer approach, but the types of benefits are too numerous and their phenomenon too
complex to pursue primary research given the financial and time constraints of this
project.
The general steps taken to estimate non-market benefits of air pollution abatement are as
follows:
1. Identifying Benefits
2. Quantifying Benefits
3. Monetizing Benefits
The first step involves describing a qualitative relationship between changes in pollutant
emissions and ambient concentrations, and subsequently between ambient concentrations
and environmental effects.
Identifying the benefits of air pollution abatement is equivalent to identifying the
damages that are reduced or avoided. These damages fall into three broad categories
(adopted from Freeman, 1993):
1. Direct damages to humans
2. Indirect damages to humans through ecosystems
3. Indirect damages to humans through nonliving systems
Direct damages to humans include health damages, as well as aesthetic damages such as
unpleasant odor, noise or poor visibility. Indirect damages to humans through ecosystems
consist of productivity damages in the form of crop reduction, and damages to forests and
commercial fisheries; recreation damages (lakes, rivers, etc.), and intrinsic or nonuse
damages. The latter are damages to ecological resources that are not motivated by
people’s own use of these resources. For example, people value endangered species or
rare ecosystems, even though they do not have the intention to ever see or experience
them. Finally, indirect damages to humans that occur through nonliving things include
damages to materials and structures, such as soiling and corrosion.
The second step, quantifying benefits, involves establishing a functional relationship
between environmental effects and the reduction in air pollution. For example, such
90
quantitative relationships can be described by dose-response or concentration-response
functions. These functions describe the change in a health effect, say, asthma attacks, and
the concentration of the pollutant that causes the effect. In order to calculate the number
of cases that will be avoided, we need to establish a baseline exposure (number of people
affected, and the level of pollution they are subjected to) and the baseline number of
cases for each quantifiable health effect for each pollutant. These numbers are then
contrasted with the number of cases for each quantifiable health effect with the regulation
(RPS, in our case), to calculate the number of cases avoided as a result of the RPS.
The third step, monetizing or valuing the benefits, is typically specific to the
environmental effect under consideration. In what follows, we describe separately the
most common valuation methods used in the literature for each effect.
As part of the process of evaluating the evidence presented by this body of literature,
each study must be evaluated as to the soundness of the data, the analytic techniques and
the conclusions drawn by the authors.
Although the approach of this report is to determine whether these non-market benefits
can be monetized in the context of NJ RPS policymaking, there are other ways of
accounting for these benefits in policymaking without monetization. One method is
tradeoff analysis, and another involves a deliberation process. If tradeoff analysis were to
be applied here, then each of the non-market benefits would be quantified separately and
not combined into a dollar value. Tradeoffs between different impacts, such as costs and
illnesses due to sulfur dioxide emissions, would be evaluated. Such an analysis can also
be informed by stakeholders’ values pertaining to the relative importance of different
impacts have. If we used a deliberative process, for example one mandated by law,
various stakeholders would provide input. An outcome is considered successful if the
process is successfully completed.
91
2. CRITERIA FOR BENEFIT TRANSFER
Instead of conducting primary research, valuation studies often apply the benefit transfer
method. Benefit transfer entails using monetary values estimated in existing empirical
studies to assess the value of a quantified effect in a different study.
When is Benefit Transfer applicable?
1. When existing studies value similar effects.
2. When the context of the existing studies is highly similar.
3. When the existing studies are of high quality.
There are no universally accepted criteria for benefit transfer, but in most cases the
defensibility of the benefit estimates depends on the quality of the existing study. Hence
one criterion for benefit transfer is to select studies that were published in peer-reviewed
journals, and have been viewed highly by the professional community.
Benefit transfer may increase the inherent uncertainty surrounding environmental
benefits estimates, but for practical reasons (e.g. cost and time), benefit transfer is a
necessary component of policy analysis. In most situations, the question is not whether to
conduct benefit transfer, but how to improve benefit transfer to make it more reliable.
When one uses benefit transfer, there are three sources of variation between the original
study location/situation and the one to which the benefits estimates are transferred to that
one has to consider. The first is individual variation (x), the second is commodity
variation (y), and the third is other variation (z).
Consider the following hypothetical situation. Suppose that the state of New Jersey is
planning to upgrade a state park, and one proposal is to provide a swimming beach to the
lake. The benefit subject to valuation is recreational swimming at the lake. Suppose
researchers have identified studies from other parts of the country that value similar
benefits. The estimates of recreational swimming benefits may not accurately measure
the benefit of recreational swimming in New Jersey for three main reasons: 1. The
preferences of New Jersey residents may differ from the preferences of the participants of
the selected studies. This is the individual variation (x) between the participants of the
original study, and the population to which benefit transfer is applied. 2. The
characteristics of the benefit may also differ. For example, the benefit derived from the
availability of a clean beach for swimming may be different in Alaska as it is in New
Jersey. This is called the commodity variation (y). 3. The first two sources of variation
imply that the value of recreational swimming in New Jersey is different from the context
of the original study, while the third source of variation (z) implies that the value was
incorrectly estimated. Each type of variation has a fixed and a random component (see
table below).
92
Fixed Component
Random Component
Individual Variation
x
Commodity Variation
y
Other Variation
z
PI
HI
PC
HC
PO
HO
The fixed component of each variation includes variables that can be measured and
observed by the researchers, while random variations are not observable. Hence, we can
make benefit transfer more reliable by minimizing the fixed component of each variation.
Individual variation may be reduced, if one is able to estimate how the value of
recreational swimming varies with demographic and socioeconomic variables, such as
age, income, etc. Using the data on these observable variables, one may adjust the
estimates for the policy context. Similarly, if one identifies the relationship between the
benefits provided by the different commodities, then the second type of variation is
reduced. For example, one might conduct a small calibration survey for the policy site.
The results from the original study could then be weighted by the respondents’
attitude/experience values for the policy site to calibrate the transfer. Selecting highquality studies for the policy context minimizes the third type of variation. The main
criteria in study selection should include statistical tests of the explanatory power and
robustness of the models used in estimation. The calibration of the original estimates may
reduce the variation between the policy context and the original study, and hence enhance
the defensibility of benefit transfer.
93
3. DIRECT BENEFITS TO HUMANS – HEALTH BENEFITS
Linkages between air pollution and health effects are subtle and often difficult to
establish. However, the available literature provides strong evidence that air pollution has
adverse effects on human health. In general, the information that is needed for the
estimation of health benefits is similar to the input needed for estimating non-health
related benefits of air pollution abatement. The first step is to estimate the change in
emissions resulting from the regulatory action. The next step involves air quality
modeling to estimate the relationship between changes in emissions and changes in
ambient pollutant concentrations. Third, one must determine the population and risk, in
order to be able to estimate the health effects. Fourth, the health effects must be
monetized to conduct a cost-benefit analysis.
3.1 IDENTIFYING HEALTH BENEFITS
Health benefits resulting from reduced air pollution can be grouped into two broad
categories: mortality benefits and morbidity benefits. Mortality benefits are avoided
deaths from diseases caused by air pollution. Morbidity benefits refer to avoided cases of
non-fatal health effects. Air pollutants that have been linked to adverse health effects
include particulate matter (PM), ozone (O3), carbon monoxide (CO), sulfur dioxide
(SO2), and nitrogen dioxide (NO2). Until the mid-1980’s it was believed that ambient
pollutant concentrations did not have adverse health effects (Katsouyanni, 2003). Over
the last 15 years, a large body of epidemiological literature has been devoted to the study
of adverse health effects occurring at moderate and low pollutant concentrations. There is
now sufficient evidence to support the hypothesis that both chronic and acute health
effects can occur at ambient pollution levels. Current research focuses on the
consequence of acute and chronic air pollution exposure for excess cardiovascular and
respiratory morbidity and mortality.
Types of health studies
There are two main types of methods to establish a dose-response relationship for a
health effect:
1. Epidemiological studies (cohort studies, case control studies, occupational
epidemiology studies, cross-sectional studies)
2. Toxicological studies (mechanistic studies, animal studies, human studies)
Epidemiological studies attempt to establish a quantitative relationship between health
effect and air pollution using a sample drawn from a large population. The four most
commonly used types of epidemiological studies are cohort studies, case control studies,
occupational epidemiology studies, and cross-sectional studies.
94
Cohort studies follow a group of healthy people with different levels of exposure and
assess the impact of exposure on their health over time. The term cohort in epidemiology
refers to a collection of people that share some common characteristics, such as age or
ethnicity. The two common types of cohort studies, prospective and retrospective, both
start by identifying and enrolling subjects based upon the presence or absence of
exposure, without knowing whether the exposure resulted in any adverse impact. In a
typical prospective cohort study, individuals in a cohort are followed forward in time, for
a sufficiently long period, to track the appearance of a disease and disability. On the
contrary, in retrospective cohort studies first the cohorts are selected and then the
exposure histories of the participants are collected and studied. Cohort studies are used
because in this setting the issue of temporality is controlled, since the exposure precedes
the disease process. Besides producing more reliable estimates, prospective cohort studies
also allow us to study how multiple risk factors determine the onset and history of one or
more diseases.
Case control studies identify a group of individuals with a certain health condition, as
well as a group of subjects without the health effect (control group), and try to answer the
question why some people got ill, while the those in the control group did not. Case
control studies are less time-intensive and expensive than cohort studies, but they are also
subject to a greater estimation bias.
Occupational epidemiology studies people working in particular jobs as subjects.
Workers in certain occupations often have a higher exposure to a pollutant than the
general population, and hence it may be easier to identify a causal relationship between
exposure to a pollutant and the health effect. Occupational epidemiology studies may not
be appropriate for benefit transfer because their study populations may not represent the
general populations in terms of risk.
Cross-sectional studies analyze the relationship between a group’s (e.g. a metropolitan
area) health status and exposure status simultaneously. This study design does not allow
for changes in variables over time, and hence it may fail to uncover the true relationship
between exposure and a health outcome.
In toxicology, mechanistic studies examine how and why various disease processes occur
in response to toxicant exposures, and help establish a relationship between dose or
exposure and response. Animal studies also provide precise information about the
adverse response to a substance because the studies are controlled in a laboratory, and
animals are subjected to a wide range of exposures. One of the advantages of animal
studies is the researcher’s ability to extrapolate from the high doses in animal studies
down to the low doses often experienced in human exposure scenarios. Finally, human
studies can be used to extrapolate the response of humans at low doses to higher doses.
Long-term (chronic) health effects of air pollution have been evaluated by a large number
of cross-sectional and a few prospective-cohort studies (Dockery et al., 1993; Pope et al.,
1995; HEI, 2000; Pope et al., 2002). Prospective cohort study is the desirable design
because exposure precedes the health outcome, which is a necessary condition for
95
establishing a causal relationship between exposure and the health outcome. Moreover,
this study design is less subject to bias because exposure is evaluated before the health
status is known, and also more accurate data may be collected. Künzli and Tager (2000)
argue that cross-sectional and prospective cohort studies address different aspects of the
association between air pollution and mortality. Cross-sectional studies are only capable
of capturing mortality effects triggered by air pollution exposures that occurred shortly
before death, while prospective cohort studies capture all air pollution-related mortality
effects.
Health benefits due to particulate matter reductions
Adverse health effects of exposure to particles have been described in numerous
epidemiological studies. Health endpoints include all-cause and cause-specific mortality
and hospital admissions. Studies conducted in the United States and in other countries
have reported associations between changes in PM and changes in mortality and
morbidity, particularly among subgroups of people with respiratory or cardiovascular
diseases. However, the exact mechanisms by which PM influences human health are not
well understood. Earlier literature focused on PM greater than 10Pm in diameter, while in
the last decade the attention of researchers turned to fine particles such as PM2.5. Recent
research indicates that ultrafine particles (UF) less than 0.1Pm in diameter may play an
important role in the induction of toxic effects. Currently, however, data on UF exposure
and health effects are still limited.
Although the importance of long-term exposure to PM has been emphasized, most of the
attention in the literature has been devoted to short-term health effects. Two prominent
prospective-cohort studies of mortality effects of PM are Dockery et al. (1993) and Pope
et al. (1995). Both of these are prospective cohort studies. Unlike earlier studies, Dockery
et al. (1993) estimate the effect of air pollution on mortality while controlling for
individual risk factors. Pope et al. (1995) study the association between air pollution and
mortality using data from a large cohort drawn from many study areas. Pope et al. (2002)
is a continuation of the Pope et al. (1995), while HEI (2000) study is a reanalysis of the
original Pope et al. (1995) data. Information on data, methodology, and the results of
these four studies are summarized in Table 3.1.
Short-term, or acute effects of PM are well established for morbidity endpoints such as,
hospital admissions for respiratory and cardiovascular conditions. There is also evidence
of acute effects on respiratory function, lower respiratory symptoms, and increased
medication use by asthmatics (Katsouyanni, 2003). There are fewer studies available on
the long-term, or chronic, health effects of PM pollution. A few studies have linked an
increase in chronic bronchitis occurrence to an increase in ambient PM concentration
(Abbey et al., 1993; Schwartz, 1993; Abbey et al., 1995). Tables 3.6-3.9, at the end of
this chapter, summarize the available studies that assessed morbidity effects resulting in
chronic and minor illness, as well as hospital admissions. We selected studies that were
conducted in the United States and Canada, and found a positive association between the
health effect and air pollution.
96
97
552,138 adults in
151 U.S. metropolitan areas
HEI (2000)
319,000-590,000 adults in 51-102 U.S.
metropolitan areas, depending on the
PM measure and study period
552,138 adults in
151 U.S. metropolitan areas
Pope et al. (1995)
Pope et al. (2002)
8111 adults in six U.S. cities
Population
Study Location and
Dockery et al. (1993)
Study
Enrollment:
1982
Deaths through: 1998
Ambient pollution data: 1980
Enrollment:
1982
Ambient pollution data: 1980
Enrollment:
1982
14-to-16-year mortality followup
Study Period
Table 3.1. Prospective Cohort Studies of Mortality Due to Particulate Matter Pollution
Fine particle and sulfur oxide pollution were associated with
all-cause death, lung cancer and cardiopulmonary mortality.
Each 10Pg/m3 increase in fine PM pollution was associated
with approximately 4%, 6%, and 8% increase in the risk of
all-cause death, cardiopulmonary mortality, and lung cancer
mortality, respectively.
The original results of Dockery et al. (1993) and Pope et al.
(1995) were successfully replicated and validated. In
alternative models, estimated mortality effects increased in
the subgroup of subjects with less than high school
education. When sulfur dioxide
was included in models with fine particles or sulfate, the
associations between these pollutants (fine particles and
sulfate) and mortality diminished.
PM pollution associated with cardiopulmonary and lung
cancer mortality but not with mortality due to other causes.
Increased mortality is associated with sulfate and fine
particulate air pollution at levels commonly found in U.S.
cities.
Air pollution was positively associated with death from lung
cancer and cardiopulmonary disease but not with death from
other causes considered together. Mortality was most
strongly associated with air pollution with fine particulates,
including sulfates.
Results
Health benefits due to ozone reductions
Ozone is formed by a chemical reaction from its precursor pollutants (volatile organic
compounds (VOCs) and nitrogen oxides (NO, NO2), and nitrous oxide (N2O)) in the
presence of heat and sunlight. Ozone concentration is the highest in the summer when the
weather is hot and sunny with relatively light winds. Electric power plants are among the
main sources of precursors pollutant emissions.
Health problems are caused by tropospheric, or ground-level, ozone. Ozone is associated
with a variety of adverse health effects ranging from minor symptoms to hospital
admissions and chronic illness. Some studies have found a link between ozone and
mortality, however there is significant uncertainty about the relationship between
mortality and high ozone concentrations, partly because of the possible confounding
effect of other pollutants such as particulate matter. Table 3.2 below summarizes the most
common adverse health effects associated with ozone.
Table 3.2. Likely Ozone-related Adverse Health Effects
Adverse Health Effect
Comment
Respiratory Hospital Admissions
A large number of studies have linked ozone to hospital
admissions for pneumonia, chronic obstructive pulmonary
disease (COPD), asthma and other respiratory ailments.
Cardiovascular Hospital Admissions
There is a link between high ozone and dysrhythmias
(abnormal heartbeat patterns).
Total Respiratory ER Visits
Minor Symptoms
Asthma Attacks
Shortness of breath
Studies have also found a link between high ozone and
emergency room visits which do not result in actual
hospital admissions.
Short-term exposure to ozone has been linked to a
variety of symptoms, including cough, sore throat and
head cold.
Ozone has specifically been linked to incidence of asthma
attacks and may be linked to the development of chronic
asthma.
Ozone associated with shortness of breath in asthmatics
and non-asthmatics.
Source: Abt Associates (1999)
Mechanistic studies of ozone yield a sufficient evidence for a biologic plausibility of
respiratory-related morbidity and mortality. For a review of mechanistic studies of ozone
see Levy et al. (2001). There is evidence from human and animal exposure studied that
long-term exposure to ozone may cause a sustained decrement in lung function. There are
well-documented molecular mechanisms for acute respiratory effects of ozone, but the
98
evidence for chronic respiratory effects is limited. There is, however, increasing evidence
that high ozone can result in the development of chronic diseases. For example,
McConnell (2002) provided the first evidence suggesting that tropospheric ozone causes
the development of childhood asthma. In high ozone-concentration cities, children who
played outdoor sports were 3 to 4 times more likely to develop chronic asthma than
children who did not play sports. In low ozone-concentration cities, children playing
sports were no more likely to develop asthma than children who did not play sports.
Because indoor ozone concentrations are generally lower than ambient concentrations,
personal exposure may not be directly related to ambient concentration. Personal
exposures to ozone are influenced by air conditioning or averting behavior, such as, more
time spent indoors. When applying concentration-response functions, one must determine
the relationship between ambient ozone concentrations and personal exposures. An
understanding of any systemic differences between the study and policy region is crucial.
The importance of personal exposures to ozone led to new recent stream of research
where the analysis of health effects is stratified by some relevant personal characteristics
(e.g. insurance status) or regional characteristics (e.g. prevalence of air conditioning).
Two recent studies, Gwynn and Thurston (2001) and Nauenberg and Basu (1999) find
that insurance status as a factor in the strength of the association between ozone and
hospital admissions for asthma. Jaffe et al. (2003) adjust for insurance status in the
relationship between air pollution and emergency department (ED) visits, by looking at
asthma ED visits and ozone among Medcaid recipients. Finkelstein et al. (2000) find that
among predictors of emergency room visits for asthma is insurance status.
Health benefits due to carbon monoxide reductions
Carbon monoxide (CO) is a colorless and odorless gas produced through incomplete
combustion of carbon-based fuels. Carbon monoxide enters the bloodstream through the
lungs and reduces the delivery of oxygen to the body’s organs and tissues. The most
vulnerable to CO are those who suffer from cardiovascular disease, particularly those
with angina or peripheral vascular disease. Fetuses and young infants, children, pregnant
women, individuals with obstructive pulmonary disease such as bronchitis and
emphysema, smokers, and individuals spending a lot of time on the street working or
doing exercise are also more susceptible to CO exposure. In health studies, high CO
concentrations have been linked to hospital admissions for asthma, chronic obstructive
pulmonary disease (COPD), dsyrhythmias, ischemic heart disease, and congestive heart
failure (CHF).
Health benefits due to sulfur dioxide reductions
Sulfur dioxide is formed when fuel, containing sulfur, such as coal and oil, is burned. The
main health effects associated with exposure to high SO2 concentrations include effects
on breathing, respiratory illness, changes in pulmonary defenses, and aggravation of
cardiovascular disease. The most susceptible groups are children, the elderly, asthmatics,
99
and people with cardiovascular and chronic lung disease (such as bronchitis and
emphysema). High SO2 levels have been linked to the following endpoints: hospital
admissions for pneumonia, ischemic heart disease, and respiratory conditions; chest
tightness, shortness of breath, and wheeze.
Health benefits due to nitrogen dioxide reductions
NO2 is a suffocating, brownish gas that is formed when fuel is burned at high
temperatures. Primary sources of NO2 are motor vehicles, electric utilities and industrial
boilers. Nitrogen dioxide can irritate the lungs and lower resistance to respiratory
infections such as influenza. There is no clear evidence on the effect of short-term
exposure to NO2 on health, but frequent exposure may cause an increased incidence of
acute respiratory illness, especially in children. NO2 has been linked to hospital
admissions for respiratory conditions, pneumonia, congestive heart failure, and ischemic
heart disease. Epidemiological studies found that NO2 has a modifying effect on PM: the
increase in mortality due to PM was found to be higher in cities where long-term NO2
concentrations were higher (Katsouyanni, 2003). In addition NO2 may have other indirect
adverse effects, as it contributes to ozone formation. Therefore the importance of NO2 for
health comes from its role as an O3 precursor and a contributor to the formation of
secondary particles.
3.2 QUANTIFYING HEALTH BENEFITS
Health benefits are typically estimated using the damage-function (DF) method the
consists of the following steps involved:
1.
2.
3.
4.
5.
6.
7.
Determining the dose-response relationship for each health effect
Determining baseline exposure
Determining the number of baseline cases for each quantifiable health effect
Number exposed u Baseline exposure u Dose-response relationship.
Determining exposure after the regulation (for each regulatory option)
Determining the number of cases for each quantifiable effect with the regulation
Determining the number of cases avoided as a result of each regulatory option
The purpose of quantification is to determine the change in the occurrence of a health
effect (y) as a result of a change in pollutant concentrations (x) between the baseline and
the control scenario. Such relationship between, say particulate matter (PM)
concentration ('x), and the change in the health effect ('y) is described by dose-response
or concentration-response (CR) functions. Dose-response or concentration-response (CR)
functions estimate the risk (of the occurrence of a health effect) per unit of exposure to a
pollutant. The two most common functional forms for the CR relationship between air
pollutants and a health effect (e.g. mortality) used in the literature are log-linear and
linear functions. The linear relationship can be described by the following equation:
y D E x
100
Where D and E are the parameters to be estimated. From the above equation it follows
that:
'y D E 'x
Log-linear CR functions have the following general form:
y J eE x
Or, equivalently:
log( y ) D E x where D
log(J ) .
Let y1 denote the occurrence of the health effect under the baseline scenario, and let y2 be
a measure of the health effect under the control scenario. Then the relationship between
the change in PM concentration and the health effect may be written as follows:
'y
ª e E x1
º
J e E PM1 « E x 1»
2
¬e
¼
y1 y2
y1 ª¬e E 'x 1º¼
In a sufficiently large population, some people develop a disease that can be attributed to
air pollution regardless of whether they were exposed or not. Relative risk (RR) is a
measure that tells us the seriousness of exposure to a known risk factor. It is defined as
the risk is for those exposed relative to those who are not exposed. For example, if the
risk of developing a disease in the exposed population is 5%, while in the non-exposed
population it is 1%, the relative risk is 5. A high risk factor indicates strong evidence
between exposure to a pollutant and the health effect.
The risk factor associated with a log-linear CR function has the following form:
RRlog -linear
y2
y1
e E 'x
Epidemiological studies typically report the relative risk, rather than the CR-coefficients.
The above equations may be used to calculate the coefficients from the reported RRvalues.
Use of thresholds in CR functions
Typically, dose-response functions that have been estimated for health effects describe a
linear no-threshold relationship. This means that every unit of exposure contributes
equally to aggregate risk in a large population of people. For example, a linear nothreshold dose-response function treats the case of one person being exposed to one
hundred units of the pollutant, and ten people subjected to ten units of pollution equally
(simply, as 100 units of human exposure). In some cases, the use of such dose-response
functions may not be appropriate. Thresholds may be incorporated into the analysis even
when one uses CR-functions that were derived under the no-threshold assumptions.
While the possible existence of a threshold in concentration-response relationships is an
101
important scientific question, there is currently no scientific basis for selecting
appropriate threshold levels.
Health studies estimating CR relationship often attempt to justify their linearity
assumption. For example, support for the no-threshold assumption in mortality is
provided by a recent study by Vedal et al. (2003). They study the association between
daily inhalable particle concentrations and daily mortality in Vancouver, British
Columbia, where daily average PM10 and ozone concentrations have been very low
during the study period. After analyzing three years of data, they conclude that increases
in low concentrations of air pollution are associated with daily mortality.
To determine the number of baseline cases, we need to identify the segment of population
that is exposed, and the number of people exposed within each segment, as well as the
level, duration and frequency of exposure. In order to obtain accurate estimates, we also
have to control for averting behavior (that is, people with known risk may act to avoid
exposure).
Quantifying Mortality Benefits
In valuation studies mortality benefits linked to particulate matter (PM) tend to dominate
total monetized benefits of air pollution abatement. The relationship between mortality
and ambient PM concentration is well established, while this is not the case for other
pollutants. Moreover, there is some evidence that there are synergistic effects between
PM and other pollutants (e.g. ozone). Therefore it is desirable to transfer estimates from
studies that consider multiple pollutants as explanatory variables in regression models.
Quantifying Mortality due to Particulate Matter
In epidemiological studies of PM, typical measurable health endpoints include all-cause
and cause specific mortality, as well as hospital admissions and emergency room visits.
Tables 3.3 and 3.4 summarize relative risk estimates associated with PM pollution from
the four available prospective cohort studies. The Pope et al. (2002) study estimates
relative risk associated with a 10 Pg/m3 increase in PM2.5 (particulates less than 2.5 Pm in
diameter), while the HEI (2000) study consider a 25 Pg/m3 change. Therefore, the
relative risk estimates in these tables are not comparable. At the end of this chapter, in
Tables 3.13-3.45 we list dose-response functions from available studies, and for each
relative risk estimate we derived an estimate for E. The value of 100uE can be interpreted
as the percentage change in the health effect associated with a unit increase in the
pollutant.
102
Table 3.3: Pope et al. (2002) estimates of adjusted relative risk (RR) associated with a 10 Pg/m3
change in fine particles measuring less than 2.5 Pm in diameter
Cause of Mortality
Adjusted relative risk
(95% confidence interval)
All causes
Cardiopulmonary
Lung cancer
All other causes
1979-1983
1999-2000
Average
1.04
1.06
1.06
(1.01-1.08)
(1.02-1.10)
(1.02-1.11)
1.06
1.08
1.09
(1.02-1.10)
(1.02-1.14)
(1.03-1.16)
1.08
1.13
1.14
(1.01-1.16)
(1.04-1.22)
(1.04-1.23)
1.01
1.01
1.01
(0.97-1.05)
(0.97-1.06)
(0.95-1.06)
Source: Pope et al. (2002), Table 2
Relative risk is adjusted for age, sex, race, smoking, education, body mass, alcohol consumption, occupational
exposure, and diet.
Table 3.4: Relative Risks of Mortality Associated with a 24.5 Pg/m3 Increase in Fine Particles Using
Alternative Measures of PM in the Reanalysis of the Pope et al. (1995) study.
Cause of Mortality
All cause
Cardiopulmonary
Lung cancer
Median PM2.5
Pope et al. (1995)
measure
1.18
(1.09–1.26)
1.30
(1.17–1.44)
1.00
(0.79–1.28)
Adjusted relative risk
(95% confidence interval)
Median PM2.5
HEI (2000)
measure
1.14
(1.06–1.22)
1.26
(1.14–1.39)
1.08
(0.88–1.32)
Mean PM2.5
Pope et al. (2000)
measure
1.12
(1.06–1.19)
1.26
(1.16–1.38)
1.08
(0.88–1.32)
Source: HEI (2000), Summary Table 4, p.21
Relative risks were calculated for a change in the pollutant of interest equal to the difference in mean concentrations
between the most-polluted city and the least-polluted city. In the Pope et al. (1995) sudy, this difference for fine
particles was 24.5 Pg/m3.
HEI (2000) tested whether the relationship between ambient concentrations/exposure and
mortality was linear using the data of Pope et al. Support for both linear and nonlinear
relationships was found, depending upon the analytic technique used. Both Pope et al.
(1995) and Dockery et al. (1993) used the Cox proportional hazard regression model,
under which hazard functions for mortality at two pollutant levels are proportional and
invariant in time.
103
HEI (2000) evaluate the applicability of the Cox proportional hazard model, using
flexible concentration-response models, for the data used in Dockery et al. (1993) and
Pope et al. (1995). The small number of study locations in Dockery et al. (1993) afforded
only a limited opportunity to define the shape of the CR-function. No evidence was found
against the linearity of the relationship for PM. For sulfate particles, however, there was
some evidence of departures from linearity at both low and high sulfate concentrations. A
similar analysis of the Pope et al. (1995) data yielded some evidence of departure from
linearity for both PM and sulfate particles. Overall, however, the Cox proportional
hazards assumption did not appear inappropriate. Finally, Pope et al. (2002) conclude that
within the range of pollution observed in the study, the CR relationship appears to be
monotonic and nearly linear.
Quantifying Mortality due to Ozone
There is considerable epidemiologic evidence concerning the relationship between
ambient ozone concentrations and human mortality risks. Because ozone contributes to
acute (short-term) health effects, the association between daily ozone concentrations and
daily mortality is of primary interest to researchers. Table 3.5 below summarizes the
findings of a number of studies that found a statistically significant relationship between
daily mortality and daily ozone concentrations. These studies show mixed findings as to
whether there is a statistically significant association between daily ozone concentrations
and daily mortality in each of the study areas.
104
Table 3.5 Studies for daily mortality and ambient ozone
Study
Study
Location/Duration
Co-pollutants in
the model
Ito and Thurston
(1996)
Cook County, Illinois
1985-1990
PM10
Kinney et al.
(1995)
Verhoeff et al.
(1996)
Los Angeles County
1985-1990
Amsterdam,
Netherlands
1986-1992
PM10
Anderson et al.
(1997)
London, England
1987-1992
Black smoke
Seoul, South Korea
1994-1998
Montreal, Quebec
1984-1993
Seoul, South Korea
1991-1997
TSP, SO2, NO2,
CO, O3
O3, SO2, NO2,
CO, PM10, PM2.5
TSP, SO2,
Lagged NO2, CO,
O3
TSP, SO2
Kwon et al. (2001)2
Goldberg et al.
(2001)
Hong et al. (2002)3
Moolgavkar et al.
(1995)
Samet et al. (1997)
Philadelphia
1973-1988
Philadelphia
1973-1988
PM10
Black smoke
TSP, SO2, NO2
Lagged CO
O3 concentration
measure (ppb)
Average same day
and previous day 1hr maxima
Daily 1-hour
maximum
Daily 1-hour
maximum
(2-day lag)
8-hour average and
daily 1-hour max
(1-day lag)
8-hour average
Daily average,
3-day running mean
8-hour average
Daily average
2-day average
Relative risk and
95% confidence
interval for a 25
ppb increase in O3
1.016
(1.004-1.029)
1.000
(0.089-1.010)
with PM10
1.024
(0.974 -1.078)
with black smoke
1.014
(0.984 - 1.046)
1.029
(1.015 - 1.042)
1.01
(1.002-1.017)
1.033
(1.017-1.05)
1.06
(1.02-1.10)
1.015
(1.004-1.026)
1.024
(1.008-1.039)
Notes:
1. O3 was measured in µg/m3. To convert to ppb, ozone concentrations in µg/m3 were divided by 1.96. In general,
the conversion factor depends on the temperature in the study area.
2. The effect of air pollution on daily mortality of patients with congestive heart failure was studied.
3. Only ischemic stroke mortality was included in the study.
Quantifying Mortality due to Carbon Monoxide
A number of studies examined the relationship between daily mortality and
concentrations of CO. Cardiovascular mortality was found strongly associated with CO
concentrations. The table below summarizes the results and characteristics of several
studies.
105
106
CO, SO2, black
smoke
Athens, Greece
1987-1991
All ages
Touloumi et
al. (1996)
CO, SO2, and black smoke significant in
single pollutant models
CO significant in single-pollutant model,
not significant in any multi-pollutant
model.
Mortality from
natural
causes
CO, O3, PM10,
SO2, NOx
São Paulo, Brazil
1990-91
Elderly (65+ years)
Saldiva et al.
(1995)
Total
mortality
In single pollutant model, CO is
significant, and PM10 and O3 are
marginally significant. In the model with
CO and PM10, both CO and PM10 are
significant.
Nonaccidental
mortality
CO, O3, PM10
Los Angeles County
1985-1990
All ages
Kinney et al.
(1995)
Main Findings
Significant effect found in all twopollutant models. Controlling for CO,
significant effect found for SO4, TSP,
COH, PM10 and PM2.5
Endpoints
Nonaccidental
mortality
Pollutants
CO, NO2, SO2, O3,
SO4, TSP, COH,
PM10, PM2.5
Toronto, Canada
1980-1984
All ages
Location, Study
Period and Population
Burnett et al.
(1998)
Study
Table 3.6
Studies for Carbon Monoxide and Mortality
Deaths during one month
summertime heat wave were
excluded from the analysis
Magnitude of single-pollutant
CO relationship drops
modestly with the inclusion of
PM10.
Association with cardiacrelated mortality is stronger,
but CO is also significantly
related to non-cardiac
mortality. PM10 and PM2.5
estimated from SO4, TSP and
COH
Comment
107
Phoenix, Arizona
1995-1997
Seoul, South Korea
1991-1997
Hong et al.
(2002)
Location, Study
Mar et al.
(2000)
Study
Cardiovascular
mortality
Ischemic Stroke
Mortality
(ICD9:431,434;
ICD10: 161,163)
TSP, SO2,
Lagged NO2, CO,
O3
Endpoints
PM2.5, PM10, SO2,
NO2, CO
Pollutants
Table 3.6
Studies for Carbon Monoxide and Mortality (continued)
Significantly increased relative risks
were found for same-day TSP and SO2,
for NO2 and CO with a 1-day lag, and
for O3 with a 3-day lag for each
interquartile range increase in the
pollutant concentration.
Cardiovascular mortality was
significantly associated with CO, NO2,
SO2, PM2.5, PM10, PMCF and elemental
carbon.
Main Findings
Both combustion-related
pollutants and secondary
aerosols (sulfates) were
associated with cardiovascular
mortality.
Comment
Quantifying Morbidity Benefits
Quantifying morbidity benefits is more difficult for chronic (long-term) conditions than
for acute (short-term) health effects, because it requires data on exposure over a long
period of time. The most frequently used endpoints in the epidemiologic literature are
hospital admissions for various respiratory and cardiovascular illnesses, emergency
department visits, chronic diseases (e.g. chronic bronchitis), and minor health effects,
such as upper respiratory symptoms (URS), lower respiratory symptoms (LRS), asthma
attacks, shortness of breath, work loss days, minor restricted activity days, etc.
Table 3.7 below briefly describes the available studies that found an association between
chronic illness and air pollution. Table 3.8 lists the available studies for minor illness, and
Tables 3.9 and 3.10 summarize studies of hospital admissions for respiratory and
cardiovascular causes, respectively.
108
109
Zemp et al. (1999)
Xu et al. (1993)
Schwartz (1993)
Portney and Mullahy (1990)
Study
9,651 individuals ages 18-60
Eight sites in Switzerland
1991
1,576 never-smokers
Beijing, China, Survey conducted
August-September 1986
6,138 individuals ages 30-74
Nationwide sample from the
National Health and Nutrition
Examination Survey 1974-75
1,318 persons age 17-93
Nationwide Sample form the
1979 National Health Interview
Survey
Location, Study
Table 3.7
Studies for Chronic Illness and Air Pollution
TSP, PM10,
NO2, O3
TSP, SO2
TSP, SO2
O3, TSP
Pollutants
Chronic
phlegm,
chronic
cough,
breathlessness, asthma,
dyspnea in
exertion
Chronic
bronchitis,
asthma
Chronic
bronchitis,
asthma,
shortness of
breath
(dyspnea),
respiratory
illness
Sinusitis, hay
fever, AOD
Endpoints
Single-pollutant models: PM10 and NO2
significantly associated with chronic
cough or phlegm, breathlessness and
dyspnea. Similar though less significant
associations found for TSP. No
significant effect found for O3.
Chronic bronchitis significantly higher in
the community with the highest TSP
level. TSP not linked to the prevalence of
asthma.
TSP significantly related to the
prevalence of chronic bronchitis, and
marginally significant for respiratory
illness. No effect on asthma or dyspnea.
Controlling for TSP, O3 significantly
related to the initiation (or exacerbation)
of sinusitis and hay fever; no effect on
AOD. TSP not significantly related to any
endpoint, although it is marginally
significant for AOD.
Main Findings
Respiratory illness
defined as a
significant condition,
coded by an
examining physician
as ICD8 code (460519)
Comment
110
4 Utah communities
1976
Chapman et al.
(1985)
McDonnell et al.
(1999)
California
initial survey: 1977
final survey:
1987
1,868
Seventh Day Adventists
Abbey et al.
(1995)
3,091 Seventh Day Adventists
California
final survey: 1992
5.623 young adults
California
final survey:
1987
3,914 Seventh Day Adventists
Location, Study
Abbey et al.
(1993)
Study
Persistent
cough and
phlegm
SO2, SO4,
NO3, TSP
Asthma
AOD,
chronic
bronchitis,
asthma
PM2.5
O3, PM10,
SO2, SO4,
NO2
AOD,
chronic
bronchitis,
asthma
Endpoints
TSP, O3,
SO2
Pollutants
Table 3.7
Studies for Chronic Illness and Air Pollution (continued)
Single-pollutant models: O3
significantly linked to asthma cases
in males, but not females; other
pollutants not significantly linked to
new asthma cases in males or
females. Two-pollutant models
estimated for ozone with another
pollutant; little impact found on size
of ozone coefficient
Average pollution
level from 1973-1992
used. Prior to 1987,
PM10 from TSP.
PM2.5 estimated from
visibility data
PM2.5 related to new cases of
chronic bronchitis, but not to new
cases of AOD or asthma
Persistent cough and phlegm is
higher in the community with higher
SO2, SO4 and TSP concentrations
Emphysema, chronic
bronchitis, and
asthma comprise
AOD
Comment
TSP linked to new cases of AOD
and chronic bronchitis, but not to
asthma or the severity of asthma.
O3 not linked to the incidence of
new cases of any endpoint, but O3
was linked to the severity of
asthma. No effect found for SO2.
Main Findings
Table 3.8
Studies for Minor Illness
Endpoints
Acute bronchitis
Ages 8-12
Dockery et al. (1996)
Pollutants Used in Final
Model
PM2.5
Upper respiratory
symptoms (URS)
Ages 9-11
Pope et al. (1991)
PM10
Lower respiratory
symptoms (LRS)
Ages 7-14
Schwartz et al. (1994)
PM2.5
Respiratory illness
Ages 6-7
Hasselblad et al. (1992)
NO2
Any of 19 respiratory
symptoms
Ages 18-65
Krupnick et al. (1990)
O3, PM10
Moderate or worse asthma
All ages (asthmatics)
Ostro et al. (1991)
PM2.5
Asthma attacks
Whittemore and Korn
(1980)
O3, PM10
Chest tightness, shortness
of breath, or wheeze
PM2.5
Shortness of breath
Ages 7-12
(African American
asthmatics)
Ages 18-65
Linn et al.
(1987,1988,1990) and
Roger and al. (1985)
Ostro et al. (1991)
PM2.5
Ostro (1987)
PM10
Work loss days
Study Population
Study
Minor restricted activity
days (MRAD)
Ages 18-65
Ostro and Rothschild
(1989)
PM2.5, O3
Restricted activity days
Ages 18-65
Ostro (1987)
PM2.5
111
112
Ages 64+
Ages 64+
Ages 64+
All ages
Detroit, MI, 1986-1989
Minneapolis, MN, 19861989
Birmingham, AL, 1986,
1989
Toronto, Canada, 19921994
Schwartz (1994a)
Schwartz (1994b)
Schwartz (1994c)
Burnett et al. (1997)
Sample
Population
Location and
period
Study
O3, PM10
O3, PM10
asthma (493), pneumonia
(480-487), non-asthma
COPD (491-492, 494496)
pneumonia (480-487),
COPD (490-496)
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
O3, PM10
asthma (493), pneumonia
(480-486), non-asthma
COPD (491-492, 494496)
all respiratory (464466,480-486,490494,496)
Pollutants
Endpoint
Table 3.9 Description of Hospital Admissions Studies for Respiratory Illnesses
O3 linked to respiratory
admissions, PM less strongly
linked. CO, NO2, and SO2 are not
significant in two-pollutant models.
PM10 marginally significant for
pneumonia admissions. O3
significant for COPD admissions.
Both pollutants are significant for
all respiratory admissions.
No association found for asthma
admissions. Both O3 and PM10
are significant for pneumonia and
COPD admissions.
PM 10 significant and O3 not
significant for btoh pneumonia and
COPD admissions in singlepollutant models.
Main Findings
113
Ages 30+
South Coast Air Basin,
CA, 1992-1995
Linn et al. (2000)
Ages 65+
All ages
Toronto, Canada, 19801994
Moolgavkar et al. (1997)
Seattle, WA, 1987-1994
Ages 64+
Minneapolis, MN,
Birmingham, AL, 19861991
Sheppard et al. (1999)
Sample
Population
Location and
period
Study
asthma (493), COPD
CO, O3, PM10
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
Pulmonary disease associated
more with NO2 and PM10 than
CO. High O3 concentrations seem
to present less risk.
In multi-pollutant models PM2.5
and CO are most significantly
related to asthma admissions.
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
Asthma (493)
O3, CO, PM2.5-10 significant for
asthma admissions. O3, NO2,
PM2.5 chosen in stepwise
regression for respiratory infection.
O3 and PM2.5-10 significant for
COPD admissions.
asthma (493), respiratory
infection (464,466,480487,494),COPD, (490492,496)
O3 significant in multi-pollutant
model for pneumonia admissions.
No significant effect found for
COPD admissions.
Main Findings
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
Pollutants
pneumonia (480-487),
COPD (490-496),
respiratory (480-487,
490-496)
Endpoint
(ICD9 codes)
Table 3.9 Description of Hospital Admissions Studies for Respiratory Illnesses (continued)
114
Sample
Population
Ages 65+
All ages
Ages 6-12
Ages 65+
Location and
period
Cook County, IL, Los
Angeles County, CA,
Maricopa County,
AZ, 1987-1995
Chicago, Cook
County, IL, 19851994
Toronto, Canada,
1981-1993
Vancouver, Canada,
1995-1999
Moolgavkar (2000)
Zanobetti et al. (2000)
Lin et al. (2003)
Chen et al. (2004)
Study
Diagnosis of conduction disorders or
dysrhythmias increased the associations
between hospital admissions for COPD and
pneumonia and PM10. Persons with asthma
had twice the risk of PM10-related pneumonia
admissions, and persons with heart failure
had twice the risk of PM10-induced COPD
admissions
Significant acute effect of CO was found in
boys, and SO2 showed significant effects of
chronic exposure in girls. NO2 was positively
associated with admissions in both sexes. O3
was not significant.
PM measures had a positive association with
hospital admissions for COPD. The
associations were no longer significant when
NO2 was included in the models.
PM10
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
COPD (490-496
except 493),
pneumonia (480487), asthma (493),
acute bronchitis
(466), acute
respiratory illness
(460-466)
asthma (493)
COPD (490-492,
494, 496)
In Cook and Maricopa counties there was a
weak association between admissions and
O3, and no association with PM10. In Los
Angeles gaseous pollutants other than ozone
were strongly associated with COPD
admissions. PM was significant only in singlepollutant models.
Main Findings
CO, O3, PM2.5,
PM10, SO2
Pollutants
COPD (490-496)
Endpoint
Table 3.9 Description of Hospital Admissions Studies for Respiratory Illnesses (continued)
115
Ages 64+
All ages
Toronto, Canada,
1980-1994
All ages
Toronto, Canada,
1992-1994
Tucson AZ, 19881990
Ages 64+
Detroit, MI, 19861989
Schwartz and Morris (1995)
Schwartz(1997)
Sample
Population
Location and
period
Study
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
ischemic heart
disease (410-414),
dysrhythmias (427),
congestive heart
failure (428)
NO2 and SO2 significant for ischemic heart
disease. O3, CO, and PM2.5 significant for
dysrhythmias. NO2 and CO significant for
congestive heart failure.
CO and PM10 significant in two-pollutant
models. No effect seen for other pollutants.
CO, O3, PM2.5,
PM2.5-10, PM10,
SO2
CO, O3, PM10,
SO2
Significant affect found for O3, and to a lesser
extent, for PM. Other pollutants are not
significant.
CO, O3, PM10,
SO2
Main Findings
PM10 and Co significant in two-pollutant
models for ischemic heart disease. No
significant effect found for dysrhythmias. In
two-pollutant models PM10 is significant for
congestive heart failure admissions.
Pollutants
cardiovascular
disease (390-429)
cardiac (410-414,
427-428)
ischemic heart
disease (410-414),
dysrhythmias (427),
congestive heart
failure (428)
Endpoint
(ICD9 codes)
Table 3.10 Description of Hospital Admissions Studies for Cardiovascular Illnesses
116
Sample
Population
Ages 65+
Ages 30+
All ages
All ages
Ages 65+
Location and
period
8 U.S. counties,
1988-1990
South Coast Air
Basin, CA, 19921995
Chicago, Cook
County, IL, 19851994
South Coast Air
Basin, CA, 19881995
Denver, Colorado,
1993-1997
Schwartz (1999)
Linn et al.(2000)
Mann et al. (2002)
Koken et al. (2003)
Study
CO, O3, PM10,
SO2
Pulmonary Heart
Disease, Coronary
Atherosclerosis,
Congestive Heart
Failure, Cardiac
Dysrhythmias
PM10
cardiovascular
disease (390-429),
myocardial infarction
(410), congestive
heart failure (428),
conduction disorders
(426), dysrthythmias
(427)
CO, O3, PM10
CO, O3, PM10
congestive heart
failure, myocardial
infection, cardiac
arrhythmia
ischemic heart
disease (410-414)
CO, PM10
Pollutants
cardiovascular
disease (390-429)
Endpoint
(ICD9 codes)
Main Findings
SO2 appears to be related to increased
hospital stays for cardiac dysrhythmias, and
CO is associated with congestive heart
failure. No association was found with PM or
NO2.
CO was associated with same-day IHD
admissions in persons with secondary
diagnosis of arrhythmia.
PM10-related hospital admissions for
cardiovascular diseases were almost doubled
in subjects with concurrent respiratory
infections.
CO, NO2, and to a lesser extent PM10
showed consistently significant relationship to
cardiovascular admissions. O3 was negatively
associated or not significant.
Both pollutants significant in two-pollutant
models
Table 3.10 Description of Hospital Admissions Studies for Cardiovascular Illnesses (continued)
3.3 VALUATION OF HEALTH BENEFITS
3.3.1 Monetizing mortality benefits
Environmental economics developed a number of methods for estimating health benefits
from avoided air pollution. The most popular primary methods are described below:
Contingent valuation method (CVM) is a survey-based method to determine willingnessto-pay (WTP) for a hypothetical change in environmental effects.
Averting behavior is a method to infer WTP from the costs and effectiveness of actions
taken to avoid a negative environmental effect.
Cost-of-illness (COI) or damage costs methods involve estimating direct costs (such as,
medical expenses) and indirect costs (for example, forgone earnings) of an environmental
effect.
Hedonic methods estimate WTP for an environmental amenity by inferring its value
from the market price of another (but in some sense related) good. For example, the
hedonic property values method estimates a marginal WTP function based on an
estimated relationship between housing prices and housing attributes (that include
environmental amenities, such as good visibility or air quality). In contrast, the hedonic
wages method estimates the value of environmental amenities from a worker’s WTA a
higher salary to compensate for exposure to higher levels of risk on the job.
Mortality benefits are most often monetized using a Value of Statistical Life (VSL)
estimate. VSL is a measure of the WTP for reductions in the risk of premature death
aggregated over the population experiencing the potential risk reduction. For example, if
each person out of one million people is willing to pay five dollars for a 1:1,000,000
reduction in mortality risk, then on average one life is saved, and hence the value of VSL
is $5,000,000. (EPA relies on a composite VSL estimate based on 26 VSL studies – 21 of
which use the hedonic wage method, and 5 use CVM).
Despite its widespread use, the VSL concept has been subject to criticism. One
shortcoming of the VSL is highlighted when one considers the difference between
mortality due to acute (short-term) and mortality due to chronic (long-term) exposure.
One of the basic assumptions underlying the VSL approach is that that equal increments
in fatality risk are valued equally irrespective of the initial risk. This assumption is
defensible only if the prior risk is small. Some experts have suggested that it is not
appropriate to estimate mortality that is the result of acute exposure, because people
affected usually have a pre-existing disease and a relatively high prior risk of mortality.
Alternative measures to VSL include the Value of Statistical Life Years (VSLY) lost or
saved. For example, if pollution abatement saves one person with average life expectancy
of fifty more years, then we say that the policy results in fifty life years extended. VSLY
117
can be interpreted as an age-specific VSL. Another alternative to VSL is the Quality
Adjusted Life Years (QALY) measure. QALY adjusts life-years extended for the quality
of life during those years. To estimate QALY, we need information about the path and
duration of health states, as well as we have to choose weights for the different health
states.
The primary approach of estimating VSL has been the use of hedonic wage and hedonic
price models that examine the equilibrium risk choices. The observed market decisions
(measured by wages and prices) reflect the joint influence of supply and demand in the
market. Most of the empirical literature has concentrated on valuing mortality risk by
estimating compensating differentials for on-the-job risk exposure in labor markets.
Viscusi and Aldy (2003) provide a meta-analysis of the extensive literature of VSL based
on estimates using U.S. labor market data from the last three decades. The main results of
these studies are summarized in Table 3.11. Estimates of VSL typically range from $4
million to $9 million.
The wage-risk relationship is typically estimated by regressing the observed wage of
individuals on a vector of personal characteristics, a vector of job characteristics, fatality
and non-fatality risk associated with the job, and the workers’ compensation benefits
payable for a job injury suffered by the worker. An important methodological question is
the choice of a risk measure. An ideal risk measure would reflect both the employee’s
perception of on-the-job fatality risk and the employer’s perception of such risk. Suitable
measures of the subjective risk are rarely available, and therefore the standard approach
in the literature has been the use of industry-specific or occupation-specific risk measures
(e.g. average number or rate of fatalities over a period of several years).
Hedonic wage studies vary in several aspects, such as, their labor market coverage (entire
labor force vs. specific occupations), geographic coverage (entire country vs. specific
states or regions), class of workers (e.g. blue-collar workers only), gender, union-status of
workers (e.g. union-members only), as well as, the measure of mortality risk used.
118
Table 3.11 Summary of Labor Market Studies of the Value of a Statistical Life, United States
Author (Year)
Smith (1974)
Thaler and
Rosen (1975)
Smith (1976)
Viscusi
(1978a 1979)
Brown (1980)
Viscusi (1981)
Olson (1981)
Arnould and
Nichols
(1983)
Butler (1983)
Sample
Current Population
Survey (CPS) 1967
Census of
Manufactures 1963
U.S. Census 1960
Employment and
Earnings 1963
Survey of Economic
Opportunity 1967
CPS 1967 1973
Survey of Working
Conditions 1969-1970
(SWC)
National Longitudinal
Survey of Young Men
1966-71 1973
Panel Study of
Income Dynamics
(PSID) 1976
CPS 1978
U.S. Census 1970
S.C. Workers’
Compensation Data
1940-69
Risk Variable
Mean Risk
Bureau of Labor
Statistics (BLS)
1966 1967
0.000125
Average
Income
Level (2000
US$)
$29,029
Implicit VSL
(millions
2000 US$)
Society of
Actuaries 1967
BLS 1966 1967
1970
BLS 1969
subjective risk of
job (SWC)
Society of
Actuaries 1967
0.001
$34,663
$1.0
0.0001
$31,027
$5.9
0.0001
$31,842
$5.3
0.002
$49,019
$1.9
BLS 1973-1976
0.0001
$22,618
$8.3
BLS 1973
Society of
Actuaries 1967
0.0001
0.001
$36,151
NA
$6.7
$0.5-$1.3
S.C. Workers’
Compensation
Claims Data
0.00005
$22,713
$1.3
$9.2
119
Table 3.11 Summary of Labor Market Studies of the Value of a Statistical Life, United States
(continued)
Author (Year)
Sample
Risk Variable
Mean Risk
Low and
McPheters
(1983)
International City
Management
Association 1976
(police officer wages)
Dorsey and
Walzer (1983)
Leigh and
Folsom
(1984)
Smith and
Gilbert (1984
1985)
Dillingham
and Smith
(1984)
Dillingham
(1985)
CPS May 1978
Constructed a risk
measure from
DOJ/FBI police
officers killed data
1972-75 for 72
cities
BLS 1976
PSID 1974 Quality of
Employment Survey
(QES) 1977
CPS 1978
CPS May 1979 BLS
industry data 1976
1979
QES 1977
BLS
BLS 1975
NY Workers’
Comp Data 1970
BLS 1976 NY
Workers’
Compensation
data 1970
Leigh (1987)
QES 1977 CPS 1977 BLS
Moore and
PSID 1982 BLS 1972- NIOSH National
Viscusi
1982
Traumatic
(1988a)
Occupational
Fatality (NTOF)
Survey 1980-1985
Source: Viscusi and Aldy (2003), Table 2, p.88
Implicit VSL
(millions
2000 US$)
0.0003
Average
Income
Level (2000
US$)
$33,172
0.000052
$21,636
$11.8- $12.3
0.0001
$29,038
$36,946
$10.1-$13.3
NA
NA
$0.9
0.000082
$29,707
$4.1-$8.3
0.000008
0.00014
$26,731
$1.2
$3.2-$6.8
NA
NA
$13.3
0.00005
0.00008
$24,931
$3.2
$9.4
$1.4
Half of the studies reviewed in Viscusi and Aldy (2003) provide estimates that range
form $5 million to $12 million (in 2000 dollars). Estimates below $5 million usually are
reported by studies that use the Society of Actuaries data, which contains data on workers
that self-selected themselves to jobs that are much riskier than average. On the other
hand, studies that report estimates above $12 million tend to estimate the wage-risk
relationship indirectly. Viscusi and Aldy (2003) regard the median estimate of $7 million
from the above table as the most reliable.
These values of VSL using hedonic wage methods are similar to those generated by U.S.
product market and housing market studies.
When transferring the estimates of VSL to non-labor market contexts, as is the case of
cost-benefit analysis of the RPS, we must make sure that the preferences of the study
population and the populations in the policy context are similar. Other factors that may
120
influence the transfer of VSL estimates are the age and income distribution of the study
population. In addition to finding a positive association between income and VSL,
Viscusi and Aldy (2003) also found a statistically significant relationship between unionstatus of workers and VSL.
3.3.2
Valuation of Morbidity Benefits
There are a number of health effects (endpoints) that can be quantified (U.S. EPA, 1989),
but difficult to value in monetary terms. These effects include, for example, reduced lung
function. Currently, there are no studies available on the economic valuation of changes
in lung function. One reason is that there is no clear connection between lung function
and the economic well-being of an individual. Reduced lung function is typically
associated with other symptoms, such as cough or asthma attacks, and it is unclear
whether a temporary decrement in lung function would go unnoticed without the related
symptoms, and hence it may have no economic value.
Several endpoints reported in the literature overlap with each other, for example the
various measures of restricted activity, or their definitions are not unique. Therefore, one
must be careful not to include a combination of endpoints that could lead to double
counting of benefits.
In a number of studies, it has been found that the different methods, mentioned in the
previous sections, generate systematically different estimates for morbidity effects. The
following table contains estimates of WTP and cost-of-illness for various health effects.
For each effect, both estimates are from the same source. We can conclude from these
results, that people’s WTP for avoiding certain symptoms usually exceeds the cost of that
symptom by an order of a magnitude. On the other hand there does not seem to be a
systematic relationship between the two estimates.
121
Table 3.12 Comparison of the value of morbidity effects using different valuation methods
Symptoms
WTP method
WTP
$1996
Individual COI
$1996
Berger et al. (1987)
Dollar value for one symptom day
Cough
Sinus Congestion
Throat Congestion
Itchy Eyes
Heavy Drowsiness
Headache
Nausea
All Symptoms
CVM
CVM
CVM
CVM
CVM
CVM
CVM
CVM
$114.74
$41.26
$66.34
$73.21
$214.44
$164.16
$72.30
$121.76
$18.38
$10.25
$21.55
$21.99
$2.72
$5.21
$3.78
$5.93
Chestnut et al. (1988, 1996)
Dollar value for one episode
Symptoms
Angina Episodes
Angina Episodes
Angina Episodes
Angina Episodes
WTP method
WTP
$1996
AB
CVM
CVM
CVM
Individual COI
$1996
$54.40
$57.26
$60.13
$147.45
$18.54
$18.54
$18.54
$18.54
Dickie-Gerking (1991)
Dollar value for a reduction to zero days/year of ozone
Health effects of ozone for
concentration
Over 12 pphm
Over 12 pphm
Over 12 pphm
Over 12 pphm
Over 9 pphm
Over 9 pphm
Over 9 pphm
Over 9 pphm
WTP method
WTP $1996
AB
AB
AB
AB
AB
AB
AB
AB
Medical expenses $1996
$138.53
$167.69
$249.35
$304.76
$249.35
$298.93
$380.58
$457.87
$36.45
$84.57
$67.08
$160.40
$59.79
$131.24
$94.78
$215.81
Rowe-Chestnut (1985)
Asthma Severity
Asthma Severity
Asthma Severity
Asthma Severity
WTP method
CVM
CVM
CVM
WTP
$1996
$ 631.70
$ 919.98
$ 697.86
$1996 Medical
expenses
$196.91
$196.91
$70.89
Source: U.S. EPA (2000)
122
Foregone Earnings
$1996
$116.57
$116.57
$116.57
Valuation of hospital admissions avoided
The typical approach for valuing the avoided incidence of hospital admissions is through
the use of COI method (U.S. EPA, 1999). Well-developed and detailed estimates of
hospitalization of the health effects are readily available (U.S. EPA, 2004). COI estimates
should be obtained for each health effects for which dose-response functions are
available. Valuation estimates typically have two components: cost of hospital stay, and
lost earnings due to hospitalization. As mentioned above, COI method underestimates
WTP by a factor of at least three. However, there are currently no studies available that
would estimate WTP directly.
123
CONCENTRATION-RESPONSE FUNCTIONS FOR OZONE (O3)
Table 3.13 CR-Functions for Ozone (O3)
Health Endpoint: Onset of asthma
CR-function:
y1
E
'O3
pop
=
=
=
=
ª
º
y1
'chronic asthma = «
y1 » pop
E'O3
y1
¬« (1 y1 ) e
¼»
occurrence rate of chronic asthma
O3 coefficient
change in O3 concentrations (ppm)
population sample
Summary of Estimates
Study
Location
Sample Population
Other Pollutants
McDonnell et al.
(1999)
McConnell et al.
(2002)
California
Non-asthmatic
males ages 27+
Children ages
9-16
None
Southern California
PM10, NO2
Ozone
Concentration
Measure
Annual average
8-hour O3
Annual 10:00h to
18:00h mean
ozone
concentration
E1
0.0277
(0.0135)
0.0904
(0.0213)
Notes: 1 Standard error of E in brackets.
Health Endpoint: Hospital admissions – pneumonia
CR-function:
y1
E
'O3
pop
=
=
=
=
'pneumonia admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily admissions rate for pneumonia per person
O3 coefficient
population sample
Study
Location
Sample Population
Other Pollutants
Ozone
Concentration
Measure
Daily average O3
concentration
Moolgavkar et
al.
(1997)
Schwartz
(1994b)
Minneapolis, MN
Population ages
65+
SO2, NO2, PM10
0.00370
(0.00103
Detroit, MI
Population ages
65+
PM10
Daily average O3
concentration
0.00521
(0.0013)
Schwartz
(1994c)
Minneapolis, MN
Population ages
65+
PM10
Daily average O3
concentration
0.00280
(0.00071)
Notes: 1. Standard error of E in brackets.
124
E1
Health Endpoint: Mortality
CR-function:
y1
E
'O3
pop
=
=
=
=
'mortality = ª y1 (e E 'O3 1) º pop
¬
¼
non-accidental deaths per person of any age
O3 coefficient
population sample
Study
Location
Sample Population
Other
Pollutants
Kinney et al.
(1995)
Los Angeles, CA
Population of all
ages
PM10
Moolgavkar et
al. (1995)
Ito and Thurston
(1996)
Philadelphia, PA
Population of all
ages
Population of all
ages
SO2, TSP
Population of all
ages
Population of all
ages
CO, NO2, SO2,
TSP
CO, NO2, NO,
SO2, PM2.5
Chicago, IL
Kelsall et al.
Philadelphia, PA
(1997)
Golberg et al.
Montréal, Québec
(2001)
PM10
Ozone
Concentration
Measure
Daily one-hour
maximum O3
0.00010
(0.000214)
Daily average
ozone
Daily one-hour
maximum O3
0.000611
(0.000216)
0.00068
(0.00029)
Daily average
ozone
Daily average
ozone
0.000936
(0.000312)
0.002056
(0.000475)
E1
Health Endpoint: Emergency room visits for asthma
'asthma-related ER visits=
CR-function:
BasePop
E
'O3
pop
=
=
=
=
E
BasePop
'O3 pop
baseline population
O3 coefficient
population sample
Study
Location
Sample Population
Other
Pollutants
Cody et al. (1992)
Northern New
Jersey
Northern New
Jersey
Population of all
ages
Population of all
ages
None
Population of all
ages
None
Weisel et al. (1992)
Stieb et al. (1996)
New Brunswick,
Canada
None
Ozone
Concentration
Measure
Daily five-hour
average O3
Daily five-hour
average O3
0.0203
(0.00717)
0.0443
(0.00723)
Daily one-hour
maximum O3
0.0035
(0.0018)
E1
125
Health Endpoint: Hospital admissions - all respiratory causes
'all respir. admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily admissions rate for all respiratory causes per person
O3 coefficient
population sample
CR-function:
y1
E
'O3
pop
=
=
=
=
Study
Location
Sample Population
Other Pollutants
Ozone
Concentration
Measure
Daily one-hour
maximum O3
concentrations
Daily average O3
Thurston et al.
(1994)
Toronto, Canada
Population of all
ages
PM2.5
Schwartz (1995)
New Haven, CT
PM10
Schwartz (1995)
Tacoma, WA
Population ages
65+
Population ages
65+
PM10
Daily average O3
concentrations
Burnett et al.
(1997)
Toronto, Canada
Population of all
ages
PM2.5, PM10,
NO2, SO2
Daily one-hour
maximum O3
concentrations
E1
0.00250
(9.71E-9)
0.00265
(0.00140)
0.00715
(0.00257)
0.00498
(0.00106)
Health Endpoint: Hospital admissions for chronic obstructive pulmonary disease (COPD)
CR-function:
y1
E
'O3
pop
=
=
=
=
'COPD admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily admissions rate for COPD per person
O3 coefficient
population sample
Study
Location
Sample Population
Other
Pollutants
Schwartz (1994b)
Detroit, MI
Population 65+
PM10
Population 65+
CO, NO2
Population of all
ages
CO, PM2.5,
PM10
Moolgavkar et al.
Minneapolis, MN
(1997)
Burnett et al.
Toronto, Canada
(1999)
126
Ozone
Concentration
Measure
Daily average O3
concentration
Daily average O3
concentration
Daily average O3
concentration
E1
0.00549
(0.00205)
0.00274
(0.00170)
0.00303
(0.00110)
Health Endpoint: Hospital admissions for asthma
CR-function:
y1
E
'O3
pop
=
=
=
=
'asthma admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily asthma admission rate per person
O3 coefficient
change in O3 concentrations (ppb)
sample population
Study
Location
Sample Population
Other Pollutants
Burnett et al.
(1999)
Toronto, Canada
Population of all
ages
CO, PM2.5, PM10
Ozone
Concentration
Measure
Daily average
O3
concentration
E1
0.00250
(0.000718)
Health Endpoint: Hospital admissions for respiratory infection
CR-function:
y1
E
'O3
pop
=
=
=
=
'respiratory infections admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily respiratory infections admission rate per person
O3 coefficient
sample population
Study
Location
Sample Population
Other Pollutants
Burnett et al.
(1999)
Toronto, Canada
Population of all
ages
PM2.5, NO2
Ozone
Concentration
Measure
Daily average
O3
concentration
E1
0.00198
(0.000520)
127
Health Endpoint: Hospital admissions for cardiac
CR-function:
y1
E
'O3
pop
=
=
=
=
'cardiac admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily cardiac admission rate per person
O3 coefficient
sample population
Study
Location
Burnett et al.
Toronto, Canada
(1997)
Sample Population
Other Pollutants
Population of all
ages
PM2.5, PM10
Ozone
Concentration
Measure
Daily 12-hour
average
E1
0.00531
(0.00142)
Health Endpoint: Hospital admissions for dysrhythmias
CR-function:
y1
E
'O3
pop
=
=
=
=
'cardiac admissions = ª y1 (e E 'O3 1) º pop
¬
¼
daily admission rate for dysrhythmias per person
O3 coefficient
sample population
Study
Location
Burnett et al.
Toronto, Canada
(1999)
Sample Population
Other Pollutants
Population of all
ages
CO, PM2.5
Health Endpoint: Presence of any of 19 acute respiratory symptoms (ARD)
CR-function:
*
'ARD 2 # E PM
10 'O3 pop
E
=
first derivative of the stationary probability
'O3
=
change in daily one-hour maximum O3 concentrations (ppb)
pop
=
sample population
128
Ozone
Concentration
Measure
Daily average
O3
E1
0.00168
(0.00103)
Study
Location
Sample
Population
Krupnick et al.
Glendora-Covina- Population aged
(1990)
Azusa, CA
18-65
Other Pollutants
SO2, COH
Ozone
Concentration
Measure
1-hour
maximum
E1
0.000137
(0.0000697)
Health Endpoint: Other minor health effects
Self-reported asthma
attacks
ª
º
y1
'asthma attacks = «
y1 » pop
'O3 E
y1
¬« (1-y1 ) e
¼»
daily incidence of asthma attacks = 0.027
y1
=
O3 coefficient = 0.00187
=
E
=
change in daily one-hour maximum O3
'O3
concentration (ppb)
=
population of asthmatics of all ages = 5.61% of the
pop
population of all ages
=
standard error of E = 0.000714
VE
Whittemore and
Korn (1980)
Location: Los
Angeles, CA
Other pollutants:
TSP
Respiratory and nonrespiratory conditions
resulting in minor
restricted activity
days
(MRAD)
'MRAD = ª y1 (e E 'O3 1) º pop
¬
¼
Ostro and Rothschild
(1989)
Location: U.S.
Other pollutants:
PM2.5
y1
E
=
=
'O3
pop
=
=
VE
=
daily MRAD incidence per person = 0.02137
inverse-variance weighted PM2.5 coefficient =
0.00220
change 2-week average of the daily 1-hour
maximum O3 concentration (ppb)
population 18-65
standard error of E = 0.000658
Source:
Costs and Benefits of the Clean Air Act, 1990-2010, Appendix D, Human Health Effects of Criteria Pollutants, EPA
129
CONCENTRATION-RESPONSE FUNCTIONS FOR PARTICULATE MATTER
(PM)
Table 3.25 CR-Functions for Particulate Matter (PM)
'nonaccidental mortality = - = y1 . (e –ȕ ¢ 'PM 2.5 – 1)
CR-function:
y1
=
=
E
'PM2.5 =
=
pop
? . pop
non-accidental deaths per person
PM2.5 coefficient
change in PM2.5 concentrations (ppm)
population sample
Study
Dockery et al.
(1993)
Location
6 U.S. cities
Sample Population
Population ages
25+
Other Pollutants
None
E1
0.0124
(0.00423)
Pope et al. (1995)
50 U.S. cities
Population ages
30+
None
0.006408
(0.001509)
HEI (2000)
6 U.S. cities
Population ages
25+
Air toxics
0.01327
(0.004408)
Pope et al.
(2002)
50 U.S. cities
Population ages
30+
Sulfate, SO2, NO2,
CO, O3
0.00583
(0.001963)
Health Endpoint: Hospital Admissions – obstructive lung disease
' COPD hospital admissions = - = y1 . (e –ȕ ¢ 'PM 2.5 – 1)
CR-function:
y1
=
=
E
=
'PM
=
pop
? . pop
hospital admissions for obstructive lung disease per person
PM coefficient
change in PM concentrations (ppm)
population sample
Study
Burnett et al.
(1999)
130
Location
Sample Population
Toronto, Canada
Population of all
ages
PM Metric/Other
Pollutants
PM2.5-10/O3
E1
0.0019
(0.0003)
Health Endpoint: Hospital Admissions – chronic obstructive pulmonary disease (COPD)
' COPD hospital admissions = - = y1 . (e –ȕ ¢ 'PM 2.5 – 1)
CR-function:
y1
=
=
E
=
'PM
=
pop
? . pop
hospital admissions for COPD per person
PM coefficient
population sample
Study
Schwartz (1994a)
Schwartz (1994b)
Schwartz (1994c)
Moolgavkar et al.
(1997)
Moolgavkar
(2000)
Zanobetti et al.
(2000)
Chen et al. (2004)
Location
Sample Population
Birmingham,
AL
Detroit,
MI
Minneapolis, MN
Population ages
65+
Population ages
65+
Population ages
65+
Population ages
65+
Population ages
65+
Population ages
65+
Population ages
65+
Minneapolis, MN
Los Angeles
County
Cook County, IL
Vancouver,
Canada
PM Metric/Other
Pollutants
PM10/None
PM10/O3
PM10/None
PM10/CO,O3
PM10/None
PM10
PM10/None
E1
0.00239
(0.000536)
0.00202
(0.00059)
0.00451
(0.00138)
0.00088
(0.000777)
0.00150
(0.000384)
0.007603
(0.003069)
0.01525
(0.004628)
131
Health Endpoint: Hospital Admissions – pneumonia
' pneumonia hospital admissions = - = y1 . (e –ȕ ¢ 'PM 2.5 – 1)
? . pop
CR-function:
=
y1
=
E
=
'PM
=
pop
hospital admissions for pneumonia per person
PM coefficient
population sample
Study
Location
Sample Population
PM Metric/Other
Pollutants
Schwartz (1994a)
Birmingham,
Population ages
PM10/None
AL
65+
Schwartz (1994b)
Detroit,
Population ages
PM10/O3
MI
65+
Schwartz (1994c)
Minneapolis, MN
Population ages
PM10/O3
65+
Moolgavkar et al.
Minneapolis, MN
Population ages
PM10/O3, NO2, SO2
(1997)
65+
Zanobetti et al.
10 U.S. cities
Population ages
PM10/O3, SO2, CO
(2000)
65+
E1
0.00174
(0.000838)
0.00115
(0.00039)
0.00157
(0.000677)
0.000498
(0.000505)
0.00193
(0.00039)
CONCENTRATION-RESPONSE FUNCTIONS FOR CARBON MONOXIDE
(CO)
Table 3.29 CR-Functions for Carbon Monoxide (CO)
CR-function:
y1
E
'CO
pop
=
=
=
=
'mortality
>
- y1 (e E 'CO 1)
@
Non-accidental deaths per person of any age
CO coefficient
change in CO concentrations (ppm)
population sample
132
Study
Saldiva et al. (1995)
Location
São Paulo, Brazil
Sample Population
Ages 65+
Other Pollutants
None
Touloumi et al.
(1996)
Athens, Greece
All ages
SO2
Toronto, Canada
All ages
TSP
Hong et al. (2002)
Seoul, South Korea
All ages
TSP
E1
0.015918
(0.00507)
0.005511
(0.00167)
0.0266
(0.0038)
0.0890
(0.0256)
Health Endpoint: Hospital admissions for asthma (ICD9-493)
CR-function: 'asthma admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for asthma per person
CO coefficient
change in average daily CO concentrations (ppm)
population sample
Study
Location
Sample Population
Other Pollutants
Toronto, Canada
All ages
PM2.5, PM10, O3
Sheppard
(1999)
Linn et al. (2000)
Seattle, WA
Under age 65
PM2.5
Los Angeles, CA
Population ages 30+
PM10, O3, NO2
Lin et al. (2003)
Toronto, Canada
Boys ages 6-12
NO2, SO2, O3
E1
0.0332
(0.00861)
0.0528
(0.0185)
0.0280
(0.0100)
0.1353
(0.0589)
Notes: 1.Standard error of E in brackets. 2.ICD = International Classification of Diseases
Health Endpoint: Hospital admissions for congestive heart failure
CR-function: 'cong. heart failure admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for congestive heart failure per person
CO coefficient
population sample
Study
Schwartz and Morris
(1995)
Location
Detroit, MI
Sample Population
Ages 65+
Other Pollutants
PM10
E1
0.0170
(0.00468)
Toronto, Canada
All ages
NO2
0.0340
(0.0163)
Koken et al. (2003)
Denver, CO
Ages 65+
NO2, O3, PM10
0.3328
(0.1681)
Notes: 1.Standard error of E in brackets. 2. ICD = International Classification of Diseases
133
Health Endpoint: Hospital admissions for cardiovascular disease
CR-function: 'cardiovascular admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for cardiovascular disease per person
CO coefficient
population sample
Study
Schwartz (1997)
Location
Tucson, AZ
Sample Population
Ages 65+
Other Pollutants
PM10
Schwartz (1999)
8 U.S. counties
Ages 65+
PM10
E1
0.0139
(0.00715)
0.0127
(0.00255)
Notes: 1.Standard error of E in brackets. 2. ICD = International Classification of Diseases
Health Endpoint: Hospital admissions for obstructive lung disease
CR-function: 'obst. lung disease admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for obstructive lung disease per person
CO coefficient
population sample
Study
Moolgavkar2 (1997)
Location
Minneapolis-St. Paul,
MN
Toronto, Canada
Sample Population
Population ages 65+
Population of all
ages
Notes: 1.Standard error of E in brackets. 2. Health endpoint: COPD
Other Pollutants
PM10, O3
PM2.5, PM10, O3
E1
0.0573
(0.0329)
0.0250
(0.0165)
Health Endpoint: Hospital admissions for dysrhythmias
CR-function: 'dysrhythmias admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for dysrhythmias per person
CO coefficient
population sample
Study
134
Location
Toronto, Canada
Sample Population
Population of all
Other Pollutants
PM2.5, O3
E1
0.0573
ages
(0.0229)
Health Endpoint: Hospital admissions for obstructive lung disease
CR-function: 'ischemic heart disease admissions = ª y1 (e E 'CO 1) º pop
¬
¼
y1
E
'CO
pop
=
=
=
=
daily hospital admissions rate for dysrhythmias per person
CO coefficient
population sample
Study
Location
Sample Population
Other Pollutants
Schwartz and Morris Detroit, MI
Population ages 65+ PM10
(1995)
E1
0.000467
(0.000435)
CONCENTRATION-RESPONSE FUNCTIONS FOR NITROGEN DIOXIDE
(NO2)
Table 3.36 CR-Functions for Nitrogen Dioxide (NO2)
Health Endpoint: Hospital admissions for respiratory conditions
CR-function: 'all respiratory admissions = ª y1 (e E 'NO2 1) º pop
¬
¼
y1
E
'NO2
pop
=
=
=
=
daily hospital admissions rate for respiratory conditions per person
NO2 coefficient
change in average daily NO2 concentrations (ppm)
population sample
Study
Burnett et al. (1997)2
Location
Toronto, Canada
Sample Population
Other Pollutants
E1
Population of all
PM2.5, PM10, O3
0.00378
ages
(0.00221)
Burnett et al. (1999)3 Toronto, Canada
Population of all
PM2.5, O3
0.00172
ages
(0.000521)
Notes: 1.Standard error of E in brackets. 2. Health endpoint: all respiratory, NO2 measure: 12-hour average
3.Health endpoint: respiratory infection
135
Health Endpoint: Hospital admissions for pneumonia
CR-function: 'pneumonia admissions = ª y1 (e E 'NO2 1) º pop
¬
¼
y1
E
'NO2
pop
=
=
=
=
daily hospital admissions rate for pneumonia per person
NO2 coefficient
population sample
Study
Moolgavkar (1997)
Location
Minneapolis-St. Paul,
MN
Notes: 1.Standard error of E in brackets.
Sample Population
Population ages 65+
Other Pollutants
PM10, O3, SO2
E1
0.00172
(0.00125)
Health Endpoint: Hospital admissions for congestive heart failure
CR-function: 'cong. heart failure admissions = ª y1 (e E 'NO2 1) º pop
¬
¼
y1
E
'NO2
pop
=
=
=
=
daily hospital admissions rate for congestive heart failure per person
NO2 coefficient
population sample
Study
Location
Toronto, Canada
Sample Population
Population of ages
Other Pollutants
PM2.5, O3
E1
0.00264
(0.000769)
Health Endpoint: Hospital admissions for ischemic heart disease
CR-function: 'ischemic heart disease admissions = ª y1 (e E 'NO2 1) º pop
¬
¼
y1
E
'NO2
pop
=
=
=
=
daily hospital admissions rate for ischemic heart disease per person
NO2 coefficient
population sample
Study
Location
Toronto, Canada
Sample Population
Population of ages
Other Pollutants
SO2
136
E1
0.00318
(0.000521)
CONCENTRATION-RESPONSE FUNCTIONS FOR SULFUR DIOXIDE (SO2)
Table 3.40 CR-Functions for Sulfur Dioxide (SO2)
CR-function:
y1
E
'SO2
pop
=
=
=
=
'mortality
>
- y1 (e E 'SO2 1)
@
Non-accidental deaths per person of any age
SO2 coefficient
change in SO2 concentrations (ppb)
population sample
Study
Saldiva et al. (1995)
Location
São Paulo, Brazil
Sample Population
Ages 65+
Other Pollutants
None
Touloumi et al. (1996)
Athens, Greece
All ages
CO
Hong et al. (2002)
Seoul, South Korea
All ages
TSP
E1
0.005204
(0.002229)
0.00404
(0.00006)
0.003343
(0.001126)
Health Endpoint: Emergency department (ED) admissions for cardiac
CR-function:
y1
E
'SO2
pop
=
=
=
=
'cardiac ED admissions
>
- y1 (e E 'SO2 1)
@
cardiac ED admissions per person
SO2 coefficient
population sample
Study
Stieb et al. (2000)
Location
Saint John, Canada
Sample Population
All ages
Other Pollutants
O3, PM10, PM2.5,
SO4
E1
0.00201
(0.000664)
Health Endpoint: Emergency department (ED) admissions for respiratory causes
CR-function:
y1
E
'SO2
pop
=
=
=
=
'cardiac ED admissions
>
- y1 (e E 'SO2 1)
@
respiratory ED admissions per person
SO2 coefficient
population sample
Study
Stieb et al. (2000)
Location
Saint John, Canada
Sample Population
All ages
Other Pollutants
O3, PM10, PM2.5,
SO4
E1
0.001527
(0.000460)
137
Health Endpoint: Hospital admissions for asthma
'Asthma hospital admissions
CR-function:
y1
E
'SO2
pop
=
=
=
=
>
- y 1 (e E 'SO2 1)
@
hospital admissions for asthma per person
SO2 coefficient
population sample
Study
Lin et al. (2003)
Location
Toronto, Canada
Sample Population
Children ages 6-12
Other Pollutants
PM2.5
E1
0.035266
(0.012383)
E1
0.0154
(0.0011)
Health Endpoint: Hospital admissions for COPD
CR-function:
y1
E
'SO2
pop
=
=
=
=
'COPD hospital admissions
>
- y1 (e E 'SO2 1)
@
hospital admissions for COPD per person
SO2 coefficient
population sample
Study
Moolgavkar (2000)
Location
Los Angeles County
Sample Population
Ages 0-19
Other Pollutants
None
Moolgavkar (2000)
Los Angeles County
Ages 20-64
None
Moolgavkar (2000)
Los Angeles County
Ages 65+
None
Moolgavkar (2000)
Maricopa County
Ages 65+
None
0.0125
(0.0008)
0.0113
(0.0010)
0.0138
(0.0019)
Health Endpoint: Hospital admissions for ischemic heart disease
CR-function:
y1
E
'SO2
pop
=
=
=
=
'ischemic heart disease hospital admissions
>
- y 1 (e E 'SO2 1)
@
hospital admissions for ischemic heart disease per person
SO2 coefficient
population sample
Study
138
Location
Toronto, Canada
Sample Population
All ages
Other Pollutants
None
E1
0.0009
(0.0001)
References
Abbey DE, Petersen FF, Mills PK, Beeson WL (1993). “Long-term ambient
concentrations of total suspended particulates, ozone, and sulfur dioxide and respiratory
symptoms in a non-smoking population”. Arch of Environ Health, 48(1): 33-46.
Abbey DE, Ostro B, Petersen F, Burchette RJ (1995). “Chronic respiratory symptoms
associated with estimated long-term ambient concentrations of fine particulates less than
2.5 Microns in diameter (PM2.5) and other air pollutants”. JEAEE, 5(2):137-159.
Anderson HR, Spix C, Medina S, Schouten JP, Castellsague J, Rossi G, Zmirou D,
Touloumi G, Wojtyniak B, Ponka A, Bacharova L, Schwartz J, Katsouyanni K (1997).
“Air pollution and daily admissions for chronic obstructive pulmonary disease in 6
European cities: results from the APHEA project”. Eur Respir J., 10(5):1064-71.
Abt Associates, (1999). “Adverse Health Effects Associated with Ozone in the Eastern
United States”, Washington, D.C: Clean Air Task Force, October 1999.
Arnould, R.J. and L.M. Nichols (1983). “Wage-Risk Premiums and Workers’
Compensation: A Refinement of Estimates of Compensating Wage Differential,”Journal
of Political Economy 91(2): 332-340.
Berger, Mark C. et al. (1987). "Valuing Changes in Health Risks: A Comparison of
Alternative Measures." The Southern Economic Journal 53: 977-984.
Brown, C. (1980). “Equalizing Differences in the Labor Market,” Quarterly Journal of
Economics 94(1), 113-134.
Burnett RT, Cakmak S, Brook JR, Krewski D, (1997). “The role of particulate size and
chemistry in the association between summertime ambient air pollution and
hospitalization for cardiorespiratory diseases”. Environ Health Perspect., 105(6):614-20.
Burnett R, Cakmak S, Raizenne M, Stieb D, Vincent R, Krewski D, Brook J, Philips O
and Ozkanyak H (1998). “The Association between Ambient Carbon Monoxide Levels
and Daily Mortality in Toronto, Canada”. Journal of the Air & Waste Management
Association, 48, 689-700.
Burnett RT, Smith-Doiron M, Stieb D, Cakmak S, Brook JR (1999). “Effects of
particulate and gaseous air pollution on cardiorespiratory hospitalizations”.Arch Environ
Health., 54(2):130-9.
Butler, R.J. (1983). “Wage and Injury Rate Responses to Shifting Levels of Workers'
Compensation.” In J.D. Worrall (ed.), Safety and the Workforce: Incentives and
Disincentives in Workers' Compensation. Ithaca, NY: ILR Press, pp. 61-86.
139
Chapman RS, Calafiore CC, Hasselblad V (1985). “Prevalence of persistent cough and
phlegm in young adults in relation to long-term ambient sulfur oxide exposure”. Am Rev
Respir Dis, 132:261-67.
Chen Y, Yang Q, Krewski D, Shi Y, Burnett RT, McGrail K (2004). “Influence of
relatively low level of particulate ar pollution on hospitalization for COPD in elderly
people”. Inhal Toxicol.; 16(1):21-5.
Chestnut, L.G. et al. (1988). “Heart Disease Patients' Averting Behavior, Costs of Illness,
and Willingness to Pay to Avoid Angina Episodes”. Final Report to the Office of Policy
Analysis, US Environmental Protection Agency.
Chestnut, L.G., L.R. Keller, W.E. Lanbert, and R.D. Rowe (1996) "Measuring Heart
Patients' Willingness to Pay for Changes in Angina Symptoms." Medical Decision
Making 16: 65-77.
Cody RP, Weisel CP, Birnbaum G, Lioy PJ (1992). “The effect of ozone associated with
summertime photochemical smog on the frequency of asthma visits to hospital
emergency departments”. Environ Res. 1992 Aug; 58(2):184-94.
Dillingham, A.E. (1985). “The Influence of Risk Variable Definition on Value-ofLifeEstimates,” Economic Inquiry 23(2), 277-294.
Dillingham, A.E. and R.S. Smith. (1984). “Union Effects on the Valuation of Fatal Risk.”
In
Dennis, B.O. (ed.), Proceedings of the Industrial Relations Research Association 36th
Annual Meeting, San Francisco, CA, December 28 – 30, 1983.Madison, WI: Industrial
Relations Research Association, pp. 270-277.
Dockery DW, Pope CA 3rd, Xu X, Spengler JD, Ware JH, Fay ME, Ferris BG Jr, Speizer
FE, (1993). “An association between air pollution and mortality in six U.S. cities.”, N
Engl J Med.; 329(24):1753-9.
Dockery DW, Cunningham J, Damokosh AI, Neas LM, Spengler JD, Koutrakis P, Ware
JH, Raizenne M, Speizer FE (1996). “Health effects of acid aerosols on North American
children: Respiratory symptoms”. Environ Health Perspect ; 104(5): 500-505.
Dorsey S. and N. Walzer. (1983). “Workers' Compensation, Job Hazards, and
Wages,”Industrial and Labor Relations Review 36(4), 642-654.
Finkelstein JA, Barton MB, Donahue JG, Algatt-Bergstrom P, Markson LE, Platt R,
(2000). “Comparing asthma care for Medicaid and non-Medicaid children in a health
maintenance organization”. Arch Pediatr Adolesc Med.; 154(6): 563-8.
140
Goldberg MS, Burnett RT, Brook J, Bailar JC 3rd, Valois MF, Vincent R (2001).
“Associations between daily cause-specific mortality and concentrations of ground-level
ozone in Montreal, Quebec”. Am J Epidemiol.; 154(9):817-26.
Gwynn RC, Thurston GD (2001). “The burden of air pollution: impacts among racial
minorities”.
Environ Health Perspect.; 109 Suppl 4:501-6.
Hasselblad V, Eddy DM, Kotchmar DJ. (1992). “Synthesis of environmental evidence:
nitrogen dioxide epidemiology studies”. J Air Waste Manage Assoc.; 42(5):662-71.
HEI (2000): Krewski, D., R. Burnett, M. Goldberg, K. Hoover, J. Siemiatycki, M. Jerrett,
M. Abrahamowicz and M. White. “Reanalysis of the Harvard Six Cities Study and the
American Cancer Society Study of Particulate Air Pollution and Mortality”. Health
Effects Institute. Cambridge, MA.
Hong YC, Lee JT, Kim H, Kwon HJ (2002) “Air pollution: a new risk factor in ischemic
stroke mortality”. Stroke.; 33(9):2165-9.
Ito, K. and Thurston, G. (1996). “Daily PM10/mortality associations: An investigation of
at-risk subpopulations”. J. Expos. Anal. Environ. Epidemiol. 6:79-95.
Jaffe DH, Singer ME, Rimm AA (2003). “Air pollution and emergency department visits
for asthma among Ohio Medicaid recipients, 1991-1996”. Environ Res.; 91(1):21-8.
Katsouyanni K. (2003). “Ambient air pollution and health”.Br Med Bull. 2003; 68:14356.
Kelsall JE, Samet JM, Zeger SL, Xu J (1997). “Air pollution and mortality in
Philadelphia, 1974-1988”. Am J Epidemiol.; 146(9):750-62.
Kinney, P.L., Ito, K., Thurston, G.D. (1995). “A sensitivity analysis of mortality/PM10
association in Los Angeles”, Inhalation Toxicol. 7, 59-69.
Koken PJ, Piver WT, Ye F, Elixhauser A, Olsen LM, Portier CJ. (2003). “Temperature,
air pollution, and hospitalization for cardiovascular diseases among elderly people in
Denver”. Environ Health Perspect.; 111(10):1312-7.
Künzli N, Tager IB (2000). “Long-term health effects of particulate and other ambient air
pollution: research can progress faster if we want it to”. Environ Health Perspect;
108(10):915-8.
Krupnick, A.J., Harrington, W. & Ostro, B. (1990). “Ambient Ozone and Acute Health
Effects: Evidence from Daily Data”. Journal of Environmental Economics and
Management, 18: 1-18
141
Kwon HJ, Cho SH, Nyberg F, Pershagen G (2001). “Effects of ambient air pollution on
daily mortality in a cohort of patients with congestive heart failure”. Epidemiology;
12(4):413-9.
Leigh, J.P. (1987). “Gender, Firm Size, Industry, and Estimates of the Value of
Life”,Journal of Health Economics 1, 331-344.
Leigh, J.P. and R.N. Folsom. (1984). “Estimates of the Value of Accident Avoidance at
the Job Depend on the Concavity of the Equalizing Differences Curve,” Quarterly
Review of Economics and Business 24(1), 56-66.
Lin M, Chen Y, Burnett RT, Villeneuve PJ, Krewski D (2003). “Effect of short-term
exposure to gaseous pollution on asthma hospitalisation in children: a bi-directional casecrossover analysis”.
J Epidemiol Community Health; 57(1):50-5.
Linn WS, Avol EL, Peng RC, Shamoo DA, Hackney JD (1987). “Replicated doseresponse study of sulfur dioxide effects in normal, atopic, and asthmatic volunteers”. Am
Rev Respir Dis; 136(5):1127-34.
Linn WS, Avol EL, Shamoo DA, Peng RC, Spier CE, Smith MN, Hackney JD (1988).
“Effect of metaproterenol sulfate on mild asthmatics' response to sulfur dioxide exposure
and exercise”. Arch Environ Health; 43(6):399-406.
Linn WS, Shamoo DA, Peng RC, Clark KW, Avol EL, Hackney JD (1990). “Responses
to sulfur dioxide and exercise by medication-dependent asthmatics: effect of varying
medication levels”. Arch Environ Health; 45(1):24-30.
Linn WS, Szlachcic Y, Gong H Jr, Kinney PL, Berhane KT, (2000). “Air pollution and
daily hospital admissions in metropolitan Los Angeles”. Environ Health Perspect. 2000
May;108(5):427-34.
Low, S.A. and L.R. McPheters. (1983). “Wage Differentials and the Risk of Death: An
Empirical Analysis,” Economic Inquiry 21, 271-280.
Mann JK, Tager IB, Lurmann F, Segal M, Quesenberry CP Jr, Lugg MM, Shan J, Van
Den Eeden SK (2002). ”Air pollution and hospital admissions for ischemic heart disease
in persons with congestive heart failure or arrhythmia”. Environ Health Perspect;
110(12):1247-52.
Mar TF, Norris GA, Koenig JQ, Larson TV (2000). “Associations between air pollution
and mortality in Phoenix, 1995-1997”. Environ Health Perspect; 108(4):347-53.
McConnell, R., et al. (2002) “Asthma in exercising children exposed to ozone: A cohort
study,” Lancet, 359:386–391.
142
McDonnell WF, Abbey DE, Nishino N, Lebowitz MD. (1999). “Long-term ambient
ozone concentration and the incidence of asthma in nonsmoking adults: the AHSMOG
Study”. Environ Research; 80:110-121
Moolgavkar, S.H., Luebeck, E.G., Hall, T.A. and Anderson, E.L. (1995). “Air pollution
and daily mortality in Philadelphia”. Epidemiology 6, 476-484.
Moolgavkar SH, Luebeck EG, Anderson EL (1997). “Air pollution and hospital
admissions for respiratory causes in Minneapolis-St. Paul and Birmingham”.
Epidemiology; 8(4):364-70.
Moolgavkar SH (2000). “Air pollution and hospital admissions for chronic obstructive
pulmonary disease in three metropolitan areas in the United States”. Inhal Toxicol; 12
Suppl 4:75-90.
Moore, M.J. and W.K. Viscusi. (1988a). “Doubling the Estimated Value of Life: Results
Using New Occupational Fatality Data,” Journal of Policy Analysis and Management
7(3), 476-490.
Nauenberg E, Basu K (1999). “Effect of insurance coverage on the relationship between
asthma hospitalizations and exposure to air pollution”. Public Health Rep; 114(2):135-48.
Olson, C.A. (1981). “An Analysis of Wage Differentials Received by Workers on
Dangerous Jobs,” Journal of Human Resources 16(2), 167-185.
Ostro, B.D. 1987. Air pollution and morbidity revisited: A speciation test. J of Environ
Economics Management. 14:87-98.
Ostro BD, Lipsett MJ, Wiener MB, Selner JC (1991). “Asthmatic responses to airborne
acid aerosols.” Am J Public Health.; 81(6):694-702.
Ostro BD, Rothschild S (1989). “Air pollution and acute respiratory morbidity: an
observational study of multiple pollutants”. Environ Res.; 50(2):238-47.
Pope CA 3rd, Dockery DW, Spengler JD, Raizenne ME (1991). “Respiratory health and
PM10 pollution. A daily time series analysis”. Am Rev Respir Dis; 144(3 Pt 1):668-74.
Pope CA 3rd, Thun MJ, Namboodiri MM, Dockery DW, Evans JS, Speizer FE, Heath
CW Jr (1995), “Particulate air pollution as a predictor of mortality in a prospective study
of U.S. adults.” American Journal of Respir Crit Care Med, 151(3 Pt 1):669-74.
Pope, C.A., 3rd, R.T. Burnett, M.J. Thun, E.E. Calle, D. Krewski, K. Ito and G.D.
Thurston. (2002). “Lung cancer, cardiopulmonary mortality, and long-term exposure to
fine particulate air pollution”. Jama. Vol. 287(9): 1132-41.
143
Portney, PR, and J. Mullahy (1990). "Urban Air Quality and Chronic Respiratory
Disease." Regional Science and Urban Economics 20:407-418.
Roger LJ, Kehrl HR, Hazucha M, Horstman DH (1985). “Bronchoconstriction in
asthmatics exposed to sulfur dioxide during repeated exercise”. J Appl Physiol.;
59(3):784-91.
Rowe, R.D. and L.G. Chestnut (1985). “Oxidants and Asthmatics in Los Angeles: A
Benefits Analysis”. Report to the Office of Policy Analysis, US Environmental
Protection Agency. Energy and Resource Consultants, Inc., Boulder, CO.
Saldiva PHN, Dockery DW, Pope CA, et al. (1995). “Air pollution and mortality in
elderly people: a time-series study in Saõ Paolo, Brazil”. Arch Environ Health; 50:15963.
Samet, J.M., Zeger, S.L., Kelsall, J.E., Xu, J. and Kalkstein, L.S. (1997). “Air pollution,
weather and mortality in Philadelphia, 1973-1988. Particulate Air Pollution and Daily
Mortality: Analyses of the Effects of Weather and Multiple Air Pollutants. The Phase IB
Report of the Particle Epidemiology Evaluation Project”. Health Effects Institute,
Cambridge, MA, pp. 1-29.
Schwartz J (1993). “Particulate air pollution and chronic respiratory disease”. Environ
Res; 62:7-13.
Schwartz J. (1994a). “Air pollution and hospital admissions for the elderly in
Birmingham, Alabama”. Am J Epidemiol. 1994 Mar 15;139(6):589-98.
Schwartz J (1994b). “Air pollution and hospital admissions for the elderly in Detroit,
Michigan”.
Am J Respir Crit Care Med. 1994 Sep;150(3):648-55.
Schwartz J (1994c). “PM10, ozone, and hospital admissions for the elderly in
Minneapolis-St. Paul, Minnesota”. Arch Environ Health. 1994 Sep-Oct;49(5):366-74.
Schwartz J (1995). “Short term fluctuations in air pollution and hospital admissions of
the elderly for respiratory disease”. Thorax.; 50(5):531-8.
Schwartz J (1997). “Air pollution and hospital admissions for cardiovascular disease in
Tucson”. Epidemiology. 1997 Jul;8(4):371-7.
Schwartz J (1999). “Air pollution and hospital admissions for heart disease in eight U.S.
counties”. Epidemiology. 1999 Jan;10(1):17-22.
Schwartz J, Dockery DW, Neas LM, Wypij D, Ware JH, Spengler JD, Koutrakis P,
Speizer FE, Ferris BG Jr. (1994). “Acute effects of summer air pollution on respiratory
symptom reporting in children.” Am J Respir Crit Care Med.; 150(5 Pt 1):1234-42.
144
Schwartz J, Morris R (1995). “Air pollution and hospital admissions for cardiovascular
disease in Detroit, Michigan”. Am J Epidemiol.; 142(1):23-35.
Sheppard L, Levy D, Norris G, Larson TV, Koenig JQ. (1999). “Effects of ambient air
pollution on nonelderly asthma hospital admissions in Seattle, Washington, 19871994”.Epidemiology; 10(1):23-30.
Smith, R.S. (1974). “The Feasibility of an `Injury Tax’ Approach to Occupational
Safety,” Law and Contemporary Problems 38(4), 730-744.
Smith, R.S. (1976). The Occupational Safety and Health Act: Its Goals and Its
Achievements. Washington, DC: American Enterprise Institute.
Smith, V.K. and C.C.S. Gilbert. (1984). “The Implicit Valuation of Risks to Life: A
Comparative Analysis,” Economics Letters 16, 393-399.
Stieb DM, Burnett RT, Beveridge RC, Brook JR. (1996). “Association between ozone
and asthma emergency department visits in Saint John, New Brunswick”.
Canada.Environ Health Perspect.; 104(12):1354-60.
Stieb DM, Beveridge RC, Brook JR, Smith-Doiron M, Burnett RT, Dales RE, Beaulieu
S, Judek S, Mamedov A (2000). “Air pollution, aeroallergens and cardiorespiratory
emergency department visits in Saint John, Canada”. J Expo Anal Environ Epidemiol.;
10(5):461-77.
Thaler, R. and S. Rosen. (1975). “The Value of Saving a Life: Evidence from the Labor
Market.” In N.E. Terleckyj (ed.), Household Production and Consumption. New
York:Columbia University Press, pp. 265-300.
Thurston GD, Ito K, Hayes CG, Bates DV, Lippmann M (1994). “Respiratory hospital
admissions and summertime haze air pollution in Toronto, Ontario: consideration of the
role of acid aerosols”. Environ Res.; 65(2):271-90.
Toulomi G, Samoli E, Katsoyanni K (1996). “Daily mortality and "winter type" air
pollution in Athens, Greece--a time series analysis within the APHEA project”. J
Epidemiol Comm Health 1996; 50 (Suppl 1): S47-S51.
U.S. EPA (1989). “Review of the National Ambient Air Quality Standards for Ozone:
Assessment of Scientific and Technical Information”, Staff Paper, Office of Air Quality
Planning and Standards.
U.S. EPA (1999). “Final Report to Congress on Benefits and Costs of the Clean Air Act,
1990 to 2010”, EPA 410-R-99-001
http://www.epa.gov/oar/sect812/
145
U.S. EPA (2000). “Non-Cancer Health Effects Valuation, Appendix C: Case Studies”,
Subcommittee of the EPA Social Science Discussion Group,
http://www.epa.gov/OSA/spc/htm/Homeqs.htm
U.S. EPA (2004). The Cost of Illness Handbook”, prepared for the Office of Pollution
Prevention and Toxics by Abt Associates, Cambridge, MA
http://www.epa.gov/oppt/coi/
Vedal S, Brauer M, White R, Petkau J. (2003). “Air pollution and daily mortality in a city
with low levels of pollution”. Environ Health Perspect.; 111(1):45-52.
Viscusi, W.K. (1978a). “Labor Market Valuations of Life and Limb: Empirical Evidence
and Policy Implications,” Public Policy 26(3), 359-386.
Viscusi, W.K. (1979). “Employment Hazards: An Investigation of Market Performance”.
Cambridge, MA: Harvard University Press.
Viscusi, W.K. (1981). “Occupational Safety and Health Regulation: Its Impact and Policy
Alternatives.” In J.P. Crecine (ed.), Research in Public Policy Analysis and Management.
Greenwich, CT: JAI Press, vol. 2, 281-299.
Viscusi W.K., Aldy J.E. (2003). “The Value of a Statistical Life: A Critical Review of
Market Estimates Throughout the World”, NBER Working Paper #9487
Weisel CP, Cody RP, Lioy PJ. (1995). “Relationship between summertime ambient
ozone levels and emergency department visits for asthma in central New Jersey”. Environ
Health Perspect.; 103 Suppl 2:97-102.
Whittemore AS, Korn EL (1980). “Asthma and air pollution in the Los Angeles area”.
Am J Public Health. 1980 Jul; 70(7):687-96.
Xu X, Wang L. (1993). “Association of indoor and outdoor particulate level with chronic
respiratory illness”. Am Rev Respir Dis; 148:1516-22.
Zanobetti A, Schwartz J, Gold D (2000). “Are there sensitive subgroups for the effects of
airborne particles?” Environ Health Perspect.; 108(9):841-5.
Zemp E, Elsasser S, Schindler C, et al. (1999). “Long-term ambient air pollution and
respiratory symptoms in adults (SAPALDIA study)”. The SAPALDIA Team. Am J
Respir Crit Care Med; 159:1257-66.
146
4. INDIRECT BENEFITS TO HUMANS THROUGH ECOSYSTEMS
Air pollution can cause significant ecological damages. Table 4.1 below identifies the
most important direct and indirect effects of air pollution.
Table 4.1
Ecological effects of air pollution
Pollutant Class
Acidic Deposition
Major Pollutants and
Precursors
Sulfuric acid, nitric acid
Precursors:
SO2, NOx
Short-term effects
Long-term effects
Direct toxic effect to plant
leaves and aquatic
organisms
Progressive deterioration
of soil quality and
acidification of surface
waters
Saturation of terrestrial
ecosystems with nitrogen.
Progressive nitrogen
enrichment of coastal
estuaries.
Accumulation of mercury
and dioxin in the food
chain.
Alteration of ecosystemwide energy flow and
nutrient cycling.
Nitrogen Deposition
NOx
Hazardous Air Pollutants
(HAPs)
Mercury and dioxins
Direct toxic effects to
animals.
Ozone
Tropospheric ozone
Direct toxic effects to plant
leaves.
The first step in valuing ecosystem benefits of reduced air pollution is to identify
measurable endpoints. Freeman (1997) identified the following categories of ecosystem
services to humankind:
1.
2.
3.
4.
Material inputs into economic activity (fossil fuels, minerals, animals)
Life-support services (breathable air, livable climate)
Environmental amenities used for recreation
Processing of waste products discharged into the environment
Available valuation methods can measure only some of these benefits: material inputs
and the value of environmental amenities used for active recreation. The most important
ecological effects with identifiable service flow impacts are summarized in Table 4.2.
147
Table 4.2
Ecological effects with identifiable and measurable human service flows
Pollutant Class
Acidification
(H2SO4, HNO3)
Ecosystem effect
High-elevation forest acidification
resulting in dieback
Freshwater acidification resulting in
fresh water organism (e.g. fish)
population decline.
Service flow impacted
Forest esthetics
Recreational Fishing
Changes in biodiversity in terrestrial
and aquatic ecosystems
Nitrogen Saturation and
Eutrophication (NOx)
Freshwater acidification resulting in
fresh water organism (e.g. fish)
population decline.
Estuarine eutrophication causing
oxygen depletion and changes in
nutrient cycling
Changes in biodiversity in terrestrial
and aquatic ecosystems
Existence/Non-use values of
biodiversity
Recreational and commercial fishing
Existence/Non-use values of
biodiversity
Economic analyses of air pollution control have paid less attention to ecological benefits
than direct benefits to human health. There is a complex and nonlinear relationship
between ecosystem damages and air pollution, and many impacts are difficult to measure.
The most important ecological benefits of air pollution abatement include:
x
x
x
x
x
Eutrophication of estuaries associated with atmospheric nitrogen deposition
Reduced tree growth associated with ozone exposure
Acidification of freshwater bodies associated with atmospheric nitrogen and
sulfur deposition
Accumulation of toxics in freshwater bodies associated with atmospheric toxics
deposition
Aesthetic damages to forests associated with ozone and airborne toxics
Eutrophication is a condition in an aquatic ecosystem where high nutrient concentrations stimulate
blooms of algae. Increased eutrophication from nutrient enrichment due to human activities is one of the
leading problems facing some estuaries in the Mid-Atlantic region.
148
Not all ecological benefits are quantifiable or can be monetized. For that reason, in
valuation studies attention is often restricted to ecological impacts associated with service
flows to humans, rather than broad structural changes to ecosystems. The main criteria
for including service flows in valuation studies are that they must be identifiable,
quantifiable and monetizable. Table 4.3 summarizes several service flows that satisfy
these criteria.
Table 4.3
Candidate Endpoints for Quantitative Assessment
Ecological Effect
Acidification
Eutrophication
Endpoint
1. Forest Aesthetics
Dose-Response Functions
1. Not required
2. Recreational Fishing
2. Multiple available
3. Existence Value of
Biodiversity
3. Multiple Available
Multiple Pollutant Stress
2. Site-specific
3. Site-specific
1. Site-specific
Biodiversity
Toxics Deposition
Economic Model
1. Site-specific
2. None available
1. Forest Aesthetics
1. Site-specific
2.Site-specific
Biodiversity
3.None available
4. Hunting and Wildlife
Aesthetics
4.Multiple Available
4. Site-specific
1. None available
1. None available
1. Ecosystem aesthetics
and ecosystem existence
value
There are four primary methods to monetize non-health-related benefits:
1.
2.
3.
4.
Hedonic property value methods
Travel cost methods (TCM)
Expressed preference methods/Contingent valuation
Market models
The use of hedonic property value methods and contingent valuation to monetize
ecological benefits is analogous to the valuation of human health benefits. TCM exploits
observed differences between travel distance and environmental quality of recreation site
to estimate the monetary value of each site characteristic. Market Models study the
impact of changes in ecological services on both producers and consumers of market
goods that rely on these services.
149
4.1. AGRICULTURAL PRODUCTIVITY BENEFITS
Air pollution has a negative impact on agricultural productivity. Research in the area has
focused primarily on the economic impacts of tropospheric ozone, acidic deposition, and
global climate change on commercial crops. Economic assessments of the impact of air
pollution on crop losses are sometimes associated with forestry impacts. However, as
Spash (1997) points out, forestry is a multi-product production system, in which the
economic valuation of the impacts pertains to a much wider set of issues, including
biodiversity changes, reduced aesthetics and recreation services. Some of these impacts
have only non-use values, and therefore forestry damages are poorly represented in
market-related production models that are commonly used to estimate agricultural
productivity damages. Therefore, in what follows we review only the methodology
commonly used to estimate agricultural crop losses due to air pollution.
Pollutants that have been found to have a negative impact on crop yields are ozone (O3)
and its precursor pollutants (mostly NOx), and sulfur dioxide (SO2). Chart 4.1 illustrates
the various impact pathways of air pollution on agricultural crops. Of all the pollutants,
the most extensive research in recent years has been conducted on tropospheric ozone.
Ashmore (1991) concludes that although gaseous pollutants other than ozone (namely,
SO2 and NO2) may be locally important at high concentrations, they have little economic
impact on a national scale. Only minor damage to plants has been attributed to gaseous
pollutants other than ozone and to sulfate and acid deposition. The National Crop Loss
Assessment Network (NCLAN) program found statistically significant response to SO2
only in soybeans and tomatoes. Herrick and Kulp (1987) report a negligible impact of
SO2 and NO2 within the National Acid Precipitation Assessment Program (NAPAP).
Ashmore (1991) finds that barley, clover, and lucerne are especially sensitive to SO2, but
these are minor crops associated with small potential economic benefits of pollution
control.
Acidic deposition is an indirect impact pathway between air pollution and crop yields.
While gaseous pollutants affect crops directly by foliage or above ground exposure,
acidic deposition causes changes in soil chemistry due to additions of sulfur and nitrogen.
The various positive (e.g. passive fertilization) and negative effects of acidic deposition
could potentially neutralize each other, although the final outcome is highly dependent on
edaphic conditions and crop cultivar (Spash, 1997).
150
Chart 4.1. Impact pathway illustrating the effects of air pollution on agricultural crops
Source: Holland et al. (2002)
Emissions
Transport and
atmospheric chemistry
Dry deposition
Wet deposition
1.
2.
1.
2.
3.
4.
5.
6.
7.
8.
Foliar necrosis
Physiological damage
Chlorosis
Pest performance
Leaching
Growth stimulation
Climate interactions
Etc.
1.
2.
3.
4.
5.
6.
7.
1.
2.
3.
4.
5.
Growth
Biomass allocations
Appearance
Death
Soil quality
1.
2.
3.
4.
Value of produce
Land prices
Breeding costs
Soil conditioning
costs
Soil Acidification
Mobilization of
heavy metals and
nutrients
Root Damage
Leaching from foliage
Nutrient loss from soil
Nutritional balance
Climate interactions
Pest performance
Etc.
151
For example, Adams et al. (1986) included fertilization effects of nitrate deposition, and
compared them with additional expense for lime used to reduce acidity. The economic
analysis of a 50% increase in acidic deposition resulted in a net benefit.
Corn and soybean appear to be the most sensitive crops to acidic deposition. In addition,
the available research seems to suggest that most commercial crop yields are relatively
insensitive to acidic deposition on its own (Segerson, 1991). Spash (1997) concludes that
crop damages from SO2, NO2 and acid deposition combined are 5-10% of the crop
damages of ozone pollution.
Segerson (1987) has identified a number of factors that distinguish the effects of acidic
deposition from those of ozone:
x
x
x
x
Acidic deposition affects a wide range of non-market goods while ozone affects
mostly commercial crops
Acidic deposition is a dynamic pollutant that accumulates over time, while ozone
is periodic. Therefore an economic analysis of ozone exposure may be based on
short-term studies, while acidic deposition should be based upon assessing
accumulated impacts over time.
The impacts of acidic deposition are surrounded by a greater degree of uncertainty
than those of ozone.
Ozone pollution is a more localized problem than acidic deposition.
Ozone has been observed to cause significant damages in terms of crop yield losses at
current ambient levels. Furthermore, the increased frequency and duration of hot dry
weather implied by global warming will increase the concentration of tropospheric ozone
from available precursors. The Table 4.1.1 below illustrates the damages to crops from
ozone exposure
Table 4.1.1 Processes and characteristics of crop plants that may be affected by ozone
Growth
Development
Yield
Quality
Rate
Fruit set & development
Number
Appearance: size,
shape, and color
Branching
Pattern
Storage life
Mass
Flowering
Source: Jacobson (1982) p. 296, Table 14.1.
152
Texture/cooking quality,
Nutrient content, Viability
of seeds
Although ozone-induced quality degradation may be a significant part of total economic
damages, research has almost exclusively focused on estimating changes in output
resulting from air pollution. There is currently insufficient information available as to the
importance of crop quality response (Spash, 1997).
The majority of economic assessments of ozone damage to crops have been at the
regional level. Published studies have concentrated on two main regions of the U.S.,
namely, the Corn Belt (Illinois, Indiana, Iowa, Ohio and Missouri) and California.
4.1.1 QUANTIFYING AGRICULTURAL PRODUCTIVITY BENEFITS
Dose-response functions describe the relationship between ambient ozone concentrations
and changes in crop yields. There are three main approaches for deriving dose-response
relationships:
1. Foliar injury models
2. Secondary response data
3. Experimentation.
Foliar injury models assess visible injury symptoms, quality changes, and growth
responses to air pollution, and they often require making subjective judgments by the
researcher. These were the primary methods used in the early literature. Another method
of dose-response function estimation is the use cross-sectional analysis of crop yield data
via regression techniques. Data on outdoor pollutant concentrations, actual crop yields,
and other environmental factors are need for the dose-response relationship estimation.
Examples of the use of secondary response data include Leung et al. (1982) and Rowe
and Chestnut (1985). Experimental approaches to study the effects of ozone on crops
include the use of greenhouses, field chambers (open-top and closed-top), unenclosed
plots, and the pollution gradient approach.
Using experimental methods, the National Crop Loss Assessment Network (NCLAN)
developed concentration-response relationships linking ground-level ozone to leaf
damage and reduced seed size in an effort to determine the effect of ozone on crop yield.
Estimated minimum and maximum dose-response functions for six major crops are
summarized in the table below.
153
Table 4.1.2 Ozone Exposure Response Functions for Selected Crops
Ozone
Quantity
Crop
Dose-response Function
Median
Index
Experimental
Duration
(Days)
1.311
SUM06
Max
Cotton
119
1-exp - index/78
4
SUM06
Min
Cotton
119
4
SUM06
Max
Field Corn
2.816
83
3
SUM06
Min
Field Corn
4.307
83
3
SUM06
Max
2.329
85
3
SUM06
Min
2.329
85
3
SUM06
Max
Grain
Sorghum
Grain
Sorghum
Peanut
2.219
112
4
SUM06
Min
Peanut
2.219
112
4
SUM06
Max
Soybean
1-exp -index/131.4 104
3
SUM06
Min
Soybean
1-exp - index/299.7 104
3
SUM06
Max
1-exp -index/27.2 58
2
SUM06
Min
Winter
Wheat
Winter
Wheat
58
2
1-exp - index/116.8 1-exp - index/92.4 1-exp - index/94.2 1-exp - index/177.8 1-exp - index/177.8 1-exp - index/99.8 1-exp - index/99.8 1.523
1-exp - index/72.1
1.547
2.353
Median
Duration
(Months)
Source: USEPA (1999)
Originally from Lee and Hogsett (1996)
The commonly assumed form of the dose-response function is the Weibull function,
which has the following functional form:
Q
P e
§O ·
¨ 3 ¸O
© r ¹
where Q is the observed yield, P is the hypothetical yield at zero ozone exposure, O3 is
the ozone concentration (ppb) and J is the ozone concentration when the yield is 0.37 P
and O is a shape parameter. This functional form is often used in empirical analysis partly
because of its biologic plausibility.
154
4.1.2 VALUATION OF AGRICULTURAL PRODUCTIVITY BENEFITS
The traditional approach to valuing crop losses due to air pollution was to calculate
monetary equivalents of the approximated losses by multiplying decreased yields by the
current market price to give a producer benefit estimate equal to total revenue. It has been
shown that this method is likely to overestimate the gain to producers from ozone
reductions. In addition, traditional methods often ignored farmers’ reactions in terms of
changing the input mix and cross-crop substitution.
In more recent empirical work agricultural production models have been used to estimate
the economic costs associated with yield losses due to air pollution. These models
estimate the social benefit from reduced ozone damages. The social benefit from a
reduction in the ozone air pollution is the change in total surplus minus the change in
deficiency/transfer payments. The social benefit, or total surplus, consists of producer
surplus (the total difference between the market price and the willingness-to-supply on
each unit sold) and the consumer surplus (the total difference between the market price
and the willingness-to-pay on each unit bought). On one hand, a reduction in crop
damage reduces costs, and hence increases the supply of the crop. This contributes to an
increase in the producer surplus, ceteris paribus. On the other hand, altered levels of
ozone pollution may affect the attributes of a crop, thus changing the consumers’
willingness to pay, and consequently change the consumer surplus. The agricultural
production model calculates the competitive equilibrium that maximizes the total surplus
subject to resource constraints.
Various agricultural production models have been used in economic assessments of
ozone damage. Murphy et al. (1999) use the AOM8 estimate the welfare changes due to
ozone air pollution the markets for eight major crops (corn, soybeans, wheat, alfalfa hay,
cotton, grain sorghum, rice, barley). The effects of a reduction in ozone concentrations
are modeled as a shift in the production function—at lower ozone levels, more output is
obtained from a given set of inputs. AOM8 is a modified version of the agricultural
production model used in Howitt (1991). In response to output and input price changes,
AOM 8 accounts for endogenous price effects and substitution of cropping activities.
Howitt et al. (1984) studied 13 crops using the California Agriculture Resources Model
(CARM) to calculate consumer and producer surpluses. CARM is a quadratic
programming model allowed for constrained cross-crop substitution.
The U.S. EPA (1999) study used the Agricultural Simulation Model (AGSIM©, Taylor et
al., 1993). The model is able to simulate the markets (equilibrium prices and quantities)
for ozone-sensitive crops. AGSIM© is an econometric-simulation model that is based on
a large set of statistically estimated demand and supply equations for agricultural
commodities produced in the United States.
The use of an agricultural production model requires one to specify an agricultural
production function. For example, Murphy et al. (1999) use a Cobb-Douglass production
155
function, where land, water, nitrogen, and pesticides are the inputs. This specification
assumes that a given change in ozone concentrations causes the same percentage change
in output for any combination of the inputs. To estimate the changes in producer and
consumer surplus, the shift in the production function is estimated based on doseresponse functions for individual crops. In Murphy et al. (1999) the relationship between
production levels under the baseline and the policy scenario are given by the following
formula:
QiP
(1 QGAIN %i B
)Qi
100
Where QiS and QiB are production levels of crop i under the policy scenario and the
baseline scenario, respectively, and QGAIN %i is the percentage change in the yield of
crop it resulting from a reductions in ambient ozone concentrations induced by the policy.
The percentage change in the output, QGAIN %i , is estimated using dose-response
functions. Each ozone reduction scenario results in a unique set of optimal input
quantities, equilibrium output prices and quantities, and welfare measures (total,
consumer, and producer surplus).
4.2 RECREATIONAL AND COMMERCIAL FISHING BENEFITS
Recreational and commercial fishing benefits form a subgroup of ecological benefits of
air pollution. The theoretical basis for valuing ecological benefits in general, and
recreational and commercial fishing benefits in particular, is that the natural environment
provides us with services that we value (Freeman, 1997). There are no suitable methods
to comprehensively value many of these service flows (e.g. breathable air, livable
climate). Therefore in valuation studies, we are limited to valuing services flows that are
either material inputs into our economy, or provide amenities associated with marketed
services (e.g. recreation).
Three types of pollution are associated with commercial and recreational fishing:
acidification, nitrogen eutrophication, and toxics deposition. Acidification, or acid
deposition, is probably the best-studied effect of air pollution on ecosystems. The main
cause of acidification is acidic precipitation in the form of sulfuric acid (H2SO4) and
nitric acid (HNO3). These acids are formed from sulfur dioxide (SO2) and nitrogen oxides
(NOx) found in the atmosphere. Electric power plants are among the primary point
sources of SO2. On the other hand, a large share of NOx is from non-point sources
(transportation), and therefore anthropogenic NOx is more dispersed in the atmosphere
compared with anthropogenic SO2. Deposition occurs via three main pathways: (1) wet
deposition, where the pollutant is dissolved in precipitation, (2) dry deposition, which is a
direct form of deposition of gases and particles to any surface, and (3) cloud-water
deposition, when cloud or water droplets are intercepted by vegetation. Since most of the
precipitation falls on the terrestrial part of the catchments, soil properties are generally
156
strongly associated with water quality. Consequently, acidification resulting from acid
deposition usually occurs in areas with acidic soils. Throughout the world freshwater
acidification is the most serious in eastern parts of the United States and Canada, and in
Europe, particularly in Scandinavia. Reductions of anthropogenic SO2 and NOx emissions
in Europe in recent years have resulted in an improvement in acidified water bodies,
however, the same trend has not been observed in the United States (Stoddard et al.,
1999). It is believed that it may take ecosystems several decades to recover from the
impact of acidification even after emissions have been cut.
Eutrophication is the result nitrogen deposition leading to excessive nitrogen enrichment
of aquatic ecosystems, and it may adversely affect the biogeochemical cycles of
watersheds. Atmospheric nitrogen is deposited into water bodies through dry and wet
deposition. Excessive eutrophication can lead, for example, to massive algae booms,
which in turn reduces the oxygen levels and leads to habitat loss. It is estimates that 86%
of the East Coast Estuaries are susceptible to eutrophication (U.S. EPA, 1997).
Toxics deposition involves hazardous air pollutants (HAPs), as defined by the Clean Air
Act: mercury, polychlorinated biphenyls (PCBs), chlordane, dioxins, and
dichlorodiphenyltrichloro-ethane (DDT). When considering air pollution from power
plants, the most important of these pollutants is mercury. Much of mercury found in
ecosystems comes from natural sources. Mercury accumulates in fish, birds, and
mammals, and it may be dangerous to humans when the concentration exceeds a critical
level. Mercury-based statewide fish consumption advisories are fairly common in the
United States.
Table 4.2.1 below summarizes the main recreational and commercial fishing impacts
associated with acidification, eutrophication, and toxics deposition.
Table 4.2.1 Recreational and commercial fishing can be associated with the following ecological impacts of
air pollution:
Pollutant Class
Ecosystem Effect
Service Flow
Acidification
(H2SO4, HNO3)
Freshwater acidifcation resulting in fish
(and other aquatic organism) decline
Nitrogen Eutrophication
(NOx)
Toxics Deposition
(Mercury, Dioxin)
Freshwater acidifcation resulting in fish
(and other aquatic organism) decline
Aquatic bioaccumulation of mercury and
dioxin
Recreational and Commercial
Fishing
4.2.1 Quantifying Recreational and Commercial Fishing Benefits
Following emissions modeling, and transport and deposition modeling, the next step in
the process of quantifying the benefits to recreational fisheries is the use of an exposure
157
model. Unfortunately, comprehensive models to quantify the impacts of acidification,
eutrophication, and toxics deposition are currently not available.
To quantify the impact of acid deposition on fisheries, US EPA (1999) uses a regionspecific model to quantify the effects of acidification on freshwater fish populations:
Model of Acidification of Groundwater Catchments – MAGIC (Cosby et. al. (1985a,b).
MAGIC is calibrated to the watershed of an individual lake or stream and then used to
simulate the response of that system to changes in atmospheric deposition. The model
simulates the effects of acid deposition on both soils and surface waters. The simulation
typically involves seasonal or annual time steps and is implemented on decadal or
centennial time scales.
Quantifying Eutrophication
When atmospheric nitrogen is deposited in estuaries, it can lead to eutrophication.
Estimation of a dose-response relationship between nitrogen loading and water quality
changes is complicated because of the dynamic nature of ecosystems. Most likely, these
dose-response functions are nonlinear with a threshold. Unfortunately, universally
transferable dose-response functions for quantifying the effects of eutrophication have
not yet been developed. USEPA (1999) study quantified deposition-related nitrogen
loadings for three estuaries (Chesapeake Bay, Long Island Sound, Tampa Bay) using
GIS-related methods. Data on nitrogen deposition, together with information on
abatement options to reduce excess nutrient loads, was used for valuation purposes. In
addition, USEPA (1999) used specific biophysical indicators of estuarine health to
quantify the benefits. This approach is useful when there is a direct link between the
biophysical indicator and the ecological service flows. USEPA (1999) used the properties
of the seagrass bed, which provides habitat for variety organisms, and have been shown
to decline with increased nitrogen deposition, as an indicator.
Quantifying Toxics Deposition
Most damages to ecosystems are caused by five hazardous air pollutants (HAP): mercury,
PCBs, dioxins, DDT, and chlordane. The mechanism of ecosystem responses to toxic
contamination is poorly understood. Furthermore, service flow impacts of ecosystem
damages are difficult to observe because HAPs persist in aquatic ecosystems for a long
time. A comprehensive quantitative analysis with the available models and data is
currently not possible.
158
4.2.2 Valuation of Recreational and Commercial Fishing Benefits
Unlike commercial fishing, recreational fishing is to a large part a non-market activity.
Although most states charge a license fee for recreational fishing, the license fee itself is
believed not to reflect the true willingness-to-pay (WTP) for recreational fishing.
Marginal WTP for recreational fishing is a function of catch attributes (e.g. number and
the average size of fish caught) and other determinants. Environmental factors indirectly
affect WTP for recreational fishing by affecting the catch attributes. The total value of
recreational fishing to the angler can be measured by the consumer surplus, which is the
difference between WTP and the actual amount they pay or the cost they incur for the
recreational fishing day. Consumer surplus is measured by the area below the demand
curve and above the price or the cost of a recreational fishing day.
There is an extensive literature on valuation of fishing opportunities by anglers. In the
valuation literature, two primary methods have been used most often to deduce the value
of recreational fishing: travel cost method (TCM), and contingent valuation method
(CVM). TCM uses observed travel costs to visit a fishing site and per-capita visitation
rates to deduce the demand for recreational fishing. On the other hand, CVM is
questionnaire-based method, where anglers are asked hypothetical question about how
much they would be willing to pay for a day of fishing.
TCM cannot be used to measure willingness-to-accept (WTA) some degradation in an
environmental amenity (i.e. compensation demanded for an environmental damage).
Hence, TCM cannot be used to estimate the value of loss of fishing opportunity due to air
pollution. Furthermore, the use of CVM in general has generated controversy in the
valuation literature. CVM is subject to an inherent bias due to its hypothetical nature.
Study participants are often subjected to an unfamiliar market context, or they may not be
fully aware of the characteristics of the good in question, or their own budget constraints.
Some critics of CVM have pointed out that estimates of WTP obtained using CVM may
not reflect the true WTP for the non-market good, but they rather reflect the WTP pay for
moral satisfaction. The answers from CVM studies may be biased because of passive-use
motives, such as the “warm glow” effect (Andreoni, 1989). Individuals’ responses to
WTP questions serve the same function as charitable contributions, and people are
assumed to get a "warm glow" from giving. Some economists do not fully recognize
“warm glow” as an economic value.
In contrast to many earlier studies utilizing TCM or CVM, Snyder et al. (2003) use a
revealed-preference approach to estimate the value of recreational fishing. Their method
to estimate WTP for a recreational fishing day is based on observed fishing license sales.
Unlike CVM and TCM that require extensive micro-level data, Snyder et al. (2003) are
able to use aggregate, state-level data for their estimation. The following simple model
describes the methodology used in Snyder et al (2003). A representative consumer’s
utility depends on the number of recreational fishing days, X, the fishing license L, and all
other goods denoted by a composite good Z.
159
U
U ( X , L, Z )
Because X and L are complementary, we can write the following:
wU ( X , 0, Z )
wX
0
wU (0, L, Z )
wX
0
That is, the marginal utility of a fishing license or a recreational fishing day alone is zero.
The demand for annual fishing licenses can be estimated from the observable data on
license sales. By measuring the appropriate area under the demand curve, one can
estimate the average benefits per license, and consequently, the value of a recreational
fishing day.
Several problems may arise when using this method. One is that the assumed
complementarity between the fishing license and recreational fishing day holds only in
the absence of illegal fishing. A comprehensive economic model of consumer behavior
would account for illegal fishing by including the “price” of illegal fishing in the
estimation process. As Snyder et al. (2003) note, in most cases, fines are set by courts,
and therefore omitting fines should not bias the analysis, as there is no apparent
correlation between license fees and fines. Another potential problem associated with the
method of Snyder et al. (2003) is price endogeneity, that is, that causality runs not only
from price to quantity, but vice versa. To address possible price endogeneity, the authors
use the instrumental variable (IV) estimation technique, using instruments that are
exogenous to the demand for fishing licenses. The set of instruments includes variables
that are indicative bureaucratic and political proclivities of states, such as the size of the
government, and the degree or regulation and taxation.
Snyder et al. (2003) obtain estimates of the value of a recreational fishing day for 48 U.S.
states. Table 4.2.2 summarizes the results for Mid-Atlantic states that are most likely to
be affected by the New Jersey RPS.
160
Table 4.2.2 Estimates of the value of a freshwater recreational fishing day for selected states
Snyder et al. (2003)
(2000$’s)
State
Linear
IV
Semi-log
IV
Log-log
IV
cutoff
$32.8
Loglog IV
cutoff
$100
Loglog IV
cutoff
$200
Loglog IV
cutoff
$500
90%
confidence
interval based
on semi-log
New Jersey
New York
Pennsylvania
Maryland
Delaware
Notes:
1.33
1.45
2.21
1.66
1.08
1.29
1.29
3.00
1.38
0.79
0.23
0.32
2.01
0.34
0.18
0.37
0.51
3.26
0.55
0.30
0.50
0.69
4.40
0.74
0.40
0.74
1.03
6.56
1.11
0.60
0.25
0.30
0.57
0.35
0.23
i.
ii.
iii.
6.65
5.59
15.87
5.44
2.74
90%
confidence
interval based
on log-log
($200 cutoff)
0.00
3.21
0.01
3.49
2.06
7.43
0.01
3.63
0.00
2.13
Estimates are based on instrumental variables regressions on data from 1975-1989
Models of demand for three functional forms are estimated: linear, multiplicative (log-log) and
semi-log. For each functional form, two specifications are reported: one includes all relevant
substitutes of resident annual licenses, and the other includes only the price of short-term
Type 1 licenses and dummy variables for each year.
Cutoff values are used as upper limits to integrate the demand function.
A comparison of Snyder et al. (2003) estimates with the values from other studies reveals
that there is a considerable geographic variation in the estimated value of recreational
fishing. Moreover, the estimates are significantly lower than those reported in other
studies employing TCM or CVM. The differences could be due to methodological
differences, as well as, to the elimination of the biases in TCM and CVM.
161
Table 4.2.3 Comparison of Snyder et al. (2003) estimates with other studies
Study
Estimation
Method
Study Location
Type of Fishing
Valuation
Valuation Snyder et al.
(2003)i
(2000 $’s)
(2000 $’s)
King and Hof (1985)
TCM
Alabama
Trout
Miller and Hay
(1980)
Walsh et. al. (1980)
TCM
Arizona
All
CVM
Colorado
Cold water
24.57
73.14
22.01
5.86
(5.22,32.14)
2.90
(1.09, 5.14)
10.40
(10.01, 30.38
2.92
(2.80, 5.59)
11.12
(10.80, 25.18)
11.12
(10.80, 25.18)
4.43
(2.73, 6.88)
18.80
(13.34, 305.88)
4.86
(4.62, 8.06)
3.51
(2.60, 5.24)
12.97
(12.29, 26.94)
11.55
(10.37, 52.00)
Ziemer et.al.
TCM
Georgia
Warm water
27.65
(1980)
Miller and Hay
TCM
Idaho
All
56.42
(1980)
Loomis and Sorg
TCM
Idaho
Cold water
45.59
(1986)
Warm water
47.04
Miller and Hay
TCM
Maine
All
48.06
(1980)
Miller and Hay
TCM
Minnesota
All
60.60
(1980)
Haas & Weithman
TCM
Missouri
Trout
27.97
(1982)
Dutta
TCM
Ohio
Cold water
8.73
(1984)
Brown and Shalloof
TCM
Oregon
Salmon
36.47
(1984)
Steelhead
49.72
Kealy and Bishop
TCM
Wisconsin
All
51.60
(1986)
Notes:
i. Snyder et al. (2003) estimates are based on the log-log instrumental variable specification with $200 cutoff; 90%
confidence interval in brackets
Another limitation of many early studies is that they do not include a direct measure of
water quality. A notable exception is Montgomery and Needelman (1997) that consider
the special case of toxic contamination of fisheries. Toxic contamination is a special case
of pollution, because contaminants in fish become dangerous to humans eating fish
before they result in a decline in fish population. In addition, through health advisories
the public is better informed about incidences of toxic contamination than other forms of
pollution (e.g. acidification or eutrophication).
Montgomery and Needelman (1997) employ the Random Utility Model, which is a sitechoice model. A brief description of the RUM model follows. Utility associated with
recreational fishing for individual i in fishing site j is given by the following equation:
162
U ijF
VijF H ijF
where VijF is the observable portion of the utility function, and H ijF is a random error.
Standard RUM models assume that VijF is a linear function of income, M i , cost of visiting
the site, Pij , and a vector of site characteristics, X j .
VijF
E M ( M i Pij ) E X X j
where E M and E X are parameters to be estimated. Using the parameter estimates we can
estimate the inclusive value, I i , which is the maximum utility that an individual taking a
trip receives.
J
Ii
ln ¦ e
PE X X ij PE M Pij
j 1
The utility of not fishing is represented by the following equations.
U iN
Vi N H iN
Vi N
E M M i E Z Zi
Given the set of individual attributes, Z i , the probability that an individual will go fishing
on a given day is given by the following formula:
1
eP
Pri ( fish 1)
1
e
P
Ii
Ii
e E Z Zi
The economic value for an individual of improved water quality (compensating variation)
can be calculated as follows:
4 CVit
ln(e
1 ˆ2
Ii
P
e
EˆZ Z i
) ln(e
Eˆ
1 ˆ1
Ii
P
ˆ
e E Z Zi )
M
163
Table 4.2.3 Valuation studies of recreational fishing using RUM
164
Study
Morey et al. (2001)
Location
Maine rivers
Study Period
1988
Type of Fishing
Salmon fishing
Ahn et al. (2000)
1996
Trout fishing
Jakus et al. (1998)
North Carolina mountain
streams
Tennessee reservoirs
1997
Recreational fishing
Lupi and Hoehn (1998)
Great Lakes, Michigan
1994
Trout and salmon fishing
Parsons et al. (1998)
Maine lakes and rivers
1989
Pendleton and Mendelsohn
(1998)
Schumann (1998)
New England lakes
1989
North Carolina
1987-1990
Ocean Fishing
Train (1998)
Montana rivers and lakes
1987-1990
Greene et al. (1997)
Tampa Bay, Florida
1991-1992
Hoehn et al. (1997)
Michigan lakes and rivers
1994-1995
Montgomery and Needelman
(1997)
Feather et al. (1995)
New York lakes
1989
Minnesota lakes
1989
McConnell and Strand (1994)
Mid- to South-Atlantic
1987-1988
Recreational ocean
fishing
4.3 BIODIVERSITY BENEFITS
Biodiversity is a valuable environmental amenity. A number of human actions, including
anthropogenic air pollution, have led to a dramatic decline in biodiversity across the
globe (Pimm et al., 2001). Biodiversity refers not just to an accumulation of species in a
given area, but it also incorporates the ecological and evolutionary interactions among
them (Armsworth et al., 2004).
The first step in estimating biodiversity benefits is defining biodiversity. Biodiversity
encompasses four levels as it is summarized in the table below:
Table 4.3.1
Type of Biodiversity
Physical Expression
Genetic
Genes, nucleotides, chromosomes, individuals
Species
Kingdom, phyla, families, genera, subspecies,
populations
Ecosystem
Bioregions, landscapes, habitats
Functional
Ecosystem, functional, robustness ecosystem
resilience services goods
Source: Turner et al. (1999)
Genetic biodiversity is the most basic level, and it refers to the information represented in
the DNA of living organisms. Species-level biodiversity refers to the variety of species in
a given area. Because only a small fraction of the estimated 5-30 million species
currently living on the earth (Wilson, 1988) have been identified and described, empirical
estimates of the species-level biodiversity are often surrounded by a great degree of
uncertainty. Community-level biodiversity is important, because it is believed that
species-level diversity enhances the productivity and stability of ecosystems (Nunes and
van den Bergh., 2001, Odum, 1950). However, recent studies suggest that no pattern or
determinate relationship may exist between species-level diversity and stability of
ecosystems (Nunes and van den Bergh. 2001, Johnson et al. 1996). Functional diversity,
or the ecosystem’s functional robustness, refers to the ability of the ecosystem to absorb
external shocks. Unfortunately, the ecosystem’s functional diversity is still poorly
understood (Nunes and van den Bergh, 2001).
Human threats to biodiversity include activities causing habitat loss (conversion,
degradation or fragmentation) and climate change, harvesting, as well as the introduction
of exotic species that by becoming dominant competitors or effective predators may drive
many native species to extinction.
165
The electric utility sector contributes to habitat degradation (acidic deposition and
eutrophication) by emitting nitrogen oxides (NOx) and sulfur oxides (SOx), as well as to
climate change by emitting greenhouse gases into the atmosphere.
Empirical estimates of Morse et al. (1995) and Field et al. (1999) of the impact of climate
change on biodiversity illustrate the magnitude of threats to biodiversity: 4 qF average
temperature increase can reduce the number of all species in California by 5%-10%.
4.3.1 Quantifying Biodiversity Benefits
The following discussion on quantifying biodiversity benefits is based on Armsworth et
al. (2004). The traditional approach to measuring biodiversity has focuses on specieslevel biodiversity, which can be measured in two ways:
x
Richness – number of species in a given area
x
Evenness – how well distributed abundance or biomass is among species within a
community
Evenness is the greatest when species are equally abundant. For example an area that has
a total population of 100 of 10 different species, each comprising of 10 individuals, is
more diverse than a community of 1 species with 91 individuals and 9 other species with
one individual each.
A diversity index is an overall measure of diversity that usually combines aspects of
richness and evenness. One of the most commonly used diversity index is the ShannonWeiner index (H') defined below.
S
Hc ¦p
i
ln( p i )
i 1
where the summation is over all species (i.e., S is the total number of species at site), and
p i is the relative abundance of species it (i.e., p i is the number of individuals of species
divided by the total number of individuals of all species at the site). H' is high if there are
many species, or if evenness is high.
Example Calculation of H':
Suppose we study a 1-acre area in a forest and have counted 240 redbud trees, 120 post
oak trees, 40 black hickory tree, and 320 red oak trees. Species richness calculations are
summarized in the table below:
166
Species (i)
Ni
pi
ln(pi)
Redbud trees (1)
240
0.333
-1.100
-0.366
Post oak tree (2)
120
0.167
-1.792
-0.298
40
0.056
-2.882
-0.161
Red oak tree (4)
320
0.444
-0.811
-0.361
Total
720
1.000
Black hickory tree
(3)
p i ln( p i )
-1.186
Estimating relative abundances for all species can be time-consuming and difficult,
therefore species-level richness is often used as a proxy for species-level biodiversity. In
contrast to the Shannon-Weiner index this simplification places relatively large weight on
rare species. Typically, measures of species-level diversity are not applied to all species
at site, but rather to a particular taxonomic group, such as, mammals, insects, or plants
(Armsworth et al. 2004)
The choice of the spatial scale is also important, because richness increases with the size
of the area. The appropriate scale is typically an economically meaningful scale (e.g.,
individual land parcel) or an ecologically meaningful scale (e.g., habitat zone).
Once the spatial scale has been chosen, there are three aspects of biodiversity to consider
(Whittaker 1972, Schluter and Ricklefs, 1993):
x
D-diversity – the “local” diversity within each site
x
E-diversity – the change in species composition from one site to another
x
J-diversity – the “total” diversity measured over the entire suite of sites being
considered
When ecological data are not available, ecological or biological production functions may
be used to approximate the changes to biodiversity. One of the most robust and useful
ecological patterns that researchers have observed is the species-area relationship. This
relationship is often approximated by a power-law formula:
S
cA z
where c and z are positive constants.
167
This relationship describes a static pattern of biodiversity, and it is not informative about
the composition of the local community. The simplest representation of the dynamics of a
closed community takes the form:
N i
N i f i ( N1 , , N S )
for i 1, , S
S = number of interacting species, f i = per capita growth rate of each species; f i can
depend on all species’ densities. The relevant partial densities indicate whether the
interaction between any two species in the community is cooperative ( f Ni j ! 0, f Nji ! 0 ),
predatory or parasitic ( f Ni j 0, f Nji ! 0 ) or competitive ( f Ni j 0, f Nji ! 0 ). A few functional
forms of ecological production functions have been reported in the literature (e.g.
Roughgarden, 1997).
4.3.2 Valuation of Biodiversity Benefits
The monetization of biodiversity benefits requires one to assess what it is about
biodiversity that consumers value. In general, consumers’ benefit can be divided into use
value (direct such as tourism or indirect such as pollination) or non-use (intrinsic or
existence) value. Direct benefits to consumers arise in two important ways:
x
Service flows – Ecosystems provide valuable services to society, such as water
purification in natural watersheds, prevention of soil erosion and carbon
sequestration by standing forests, and recreational services such as ecotourism
and birdwatching. The service flow approach to valuation dictates that
investments in preserving or restoring biodiversity need to deliver a competitive
return relative to other investment opportunities within the economy, and hence it
does not necessarily maximize D, E or J diversity (Armsworth, 2004).\
x
Bet hedging – Conserving biodiversity provides society with a bet-hedge against
unforeseen circumstances. For example if society were overly reliant on
monocultured ecosystems, it would be vulnerable to catastrophic losses in service
provision due to disease outbreaks. Hence there is a bet-hedging benefit to
conserving J diversity.
Nonuse or existence value of biodiversity refers to the utility consumers derive from
knowing that certain species (still) exist.
Nunes and van den Bergh (2001) critically evaluate a number of biodiversity valuation
studies at each level of biodiversity value contained in Table 4.3.1. The authors conclude
that available economic valuation estimates should be regarded as providing a very
168
incomplete perspective on the value of biodiversity changes, and they provide at best the
lower bounds on that value. The tables below summarize the results form the valuation
studies that have been performed in North America. Table 4.3.2 lists the results from
contingent valuation studies estimating the individual WTP to avoid the loss of a
particular species. The main shortcoming of these single-species valuation studies is that
they do not account for species substitution and complementary effects.
Table 4.3.2 Biodiversity value estimates from single-species valuation studies
Author(s)
Study
Mean WTP estimates
(per household per year)
Stevens et al. (1997)
Restoration of the Atlantic salmon
$14.38-21.40
in one river, Massachussetts
Loomis and Larson (1994)
Conservation of the Gray Whale,
$16–18
US
Loomis and Helfand (1993)
Van Kooten (1993)
Conservation of various single
From $13 for the Sea Turtle to
species, US
$25 for the Bald Eagle
Conservation of waterfowl habitat
$50–60 (per acre)
in Canada’s wetlands region
Bower and Stoll (1988)
Conservation of the Whooping
$21–141
Crane
Boyle and Bishop (1987)
Two endangered species in
From $5 for the Striped Shiner to
Wisconsin: the Bald Eagle and
$28 for the Bald Eagle
the Striped Shiner
Brookshire et al. (1983)
Grizzly Bear and Bighorn Sheep
From $10 for the Grizzly Bear to
in Wyoming
$16 for the Bighorn Sheep
Source: Nunes and van den Bergh (2001)
Multiple-species valuation studies account for all related species, and the resulting
estimates are in general higher than those of single-species studies. Multiple-species
valuation studies are summarized in Table 4.3.3 below.
169
Table 4.3.3 Biodiversity value estimates from multiple-species valuation studies
Author(s)
Study
Mean WTP Estimates
(per household per year)
Desvousges et al. (1993)
Conservation of the migratory
$59–71
Waterfowl in the Central
Flyway
Whitehead (1993)
Conservation program for
$15
coastal nongame wildlife
Duffield and Patterson (1992)
Halstead et al. (1992)
Conservation of fisheries in
$2–4 (for residents) $12–17
Montana Rivers
(for non residents)
Preservation of the Bald Eagle,
$15
Coyote and Wild Turkey in
New England
Samples and Hollyer (1989)
Preservation of the Monk Seal
$9.6–13.8
and Humpback Whale
Hageman (1985)
Preservation of threatened and
$17.73–23.95
endangered species
populations in the US
Table 4.3.4 summarizes estimates from valuation studies that link the value of
biodiversity to the value of natural habitat conservation.
Table 4.3.4 Biodiversity value estimates from natural habitat valuation studies
Author(s)
Study
Mean WTP estimates
(per household)
Richer (1995)
Desert protection in California, US
$101
Kealy and Turner (1993)
Preservation of the aquatic system in the
$12–18
Adirondack Region, US
170
Table 4.3.4 Biodiversity value estimates from natural habitat valuation studies (continued)
Author(s)
Study
Mean WTP estimates
(per household)
Hoehn and Loomis (1993)
Diamond et al. (1993)
Enhancing wetlands and habitat in San
$96–184
Joaquin valley in California, US
(single program)
Protection of wilderness areas in
$29–66
Colorado, Idaho, Montana, and Wyoming,
US
Silberman et al. (1992)
Protection of beach ecosystems, New
$9.26–15.1
Jersey, US
Boyle (1990)
Preservation of the Illinois Beach State
$37–41
Nature Reserve, US
Loomis (1989)
Preservation of the Mono Lake,
$4–11
California, US
Smith and Desvousges (1986)
Preservation of water quality in the
$21–58 (for users)
Monongahela River Basin, US
$14–53
(for nonusers)
Mitchell and Carson (1984)
Preservation of water quality for all rivers
$242
and lakes, US
Walsh et al. (1984)
Protection of wilderness areas in
$32
Colorado, US
Table 4.3.5 summarizes estimates from valuation studies that link the value of
biodiversity to the value of natural areas with high tourism and outdoor recreation
demand.
171
Table 4.3.5 Biodiversity value estimates from ecosystem functions and services valuation studies
Author(s)
Study
Measurement
Estimates
method
Laughland et al. (1996)
Value of a water supply in
Averting behavior
Milesburg, Pennsylvania,
$14 and 36 per
household
US
Table 4.3.5 Biodiversity value estimates from ecosystem functions and services valuation studies
(continued)
Author(s)
Study
Measurement
Estimates
method
Abdalla et al. (1992)
Groundwater ecosystem
Averting behavior
$61 313–131 334
Contingent valuation
$7–22
Production function
$9.5–1.09 per acre-
in Perkasie,
Pennsylvania, US
McClelland et al. (1992)
Protection of groundwater
program, US
Torell et al. (1990)
Water in-storage on the
high plains aquifer, US
Ribaudo (1989a,b)
Water quality benefits in
foot
Averting behavior
$4.4 billion
Replacement costs
$454 million per year
Replacement costs
$35–661 million
ten regions in US
Huszar (1989)
Value of wind erosion
costs to households in
New Mexico, US
Holmes (1988)
Value of the impact of
water turbidity due to soil
annually
erosion on the water
treatment, US
Walker and Young (1986)
Value of soil erosion on
Production function
$4 and 6 per acre
(loss) agriculture revenue
in the Palouse region, US
Finally, Table 4.3.6 provides ranges of estimates for the various levels of biodiversity
value derived from the studies reviewed by Nunes and van den Bergh (2001).
172
Table 4.3.6 Summary of biodiversity values by biodiversity level
Biodiversity level
Biodiversity value type
Value ranges
Method(s) selected
Genetic and species diversity
Bioprospecting
From $175 000 to $3.2
Market contracts
million
Single species
From $5 to 126
Multiple species
From $18 to 194
Ecosystems and natural
Terrestrial habitat
From $27 to 101
habitat diversity
(non-use value)
From $9 to 51
From $8 to 96
Natural areas habitat
From $23 per trip to 23
Travel cost, tourism
(recreation)
million per year
revenues
Wetland life-support
From $0.4 to 1.2
Replacement costs
Coastal habitat (nonuse value)
Wetland habitat (nonuse value)
Ecosystems and functional
diversity
million
Soil and wind erosion
Up to $454 million per
Replacement costs,
protection
year
hedonic price,
production function
Water quality
From $35 to 661
Replacement costs,
million per year
averting expenditure
4.4 FORESTRY BENEFITS
Air pollution has been recognized as a potential problem for forests for a long time.
Sulfur dioxide, fluorides, heavy metals and ozone pose the greatest threat to forest
ecosystems. In the past, sulfur emissions, that cause acid rain, were the primary concern,
but in recent decades massive efforts to reduce this pollutant have been largely
successful. Today, in terms forestry impacts ozone may be the pollutant associated with
the greatest potential benefits.
Scientific evidence suggests that elevated tropospheric ozone levels disrupt vegetation
growth, and interfere with the respiratory function of plants carried out by photosynthesis
even at concentrations below current air quality standards (Wang et al., 1986; Reich and
173
Amudson, 1985). Sometimes ozone injury to plants has observable effects such as
yellowing or stippling of leaves, but negative impacts of ozone often occur without
accompanying visible symptoms.
Ambient ozone enters the plant through pores in the leaf or needle called stomata, where
most of the plant’s metabolic and respiratory activity occurs. Once ozone enters these
stomata, it initiates a chain reaction that destroys or damages plant proteins and enzymes,
as well as the fatty chemicals that help form cell membranes. Plants continue to suffer
damage long after the ozone exposure episode is over. Furthermore some researchers
have suggested that there are synergies between ozone and acid deposition (Hewitt,
1990).
Ozone damage to forests is a common problem in many parts of the eastern U.S.
Particularly sensitive species to ozone are the poplars (Populus spp.), white pine (Pinus
strobus), and the oak family (Quercus spp.).
Another serious threat to forest ecosystems is acid deposition in the form of nitrogen
acids due to nitrogen oxides emissions. Aluminum is naturally present in forest soils in
the form of chemical compounds that are harmless to living organisms. Nitrogen acids
cause ions of aluminum to become mobile in soil, and its toxic form, aluminum is taken
up into the tree’s roots. This may result in reduced root growth, which reduces the tree’s
ability to take up water and withstand drought. Excess nitrogen is also absorbed directly
from the air through the leaves during fog and low clouds. If ozone is present in sufficient
concentrations, exposure to this oxidant can damage the leaves, damaging respiration
processes of the organism.
Given the evidence on damage to forests and the size of the forest cover, forestry benefits
seem to play an important role in total benefits due to reduced air pollution in the
northeastern United States. According to Mid-Atlantic Integrated Assessment (MAIA) –
multi-agency effort headed by the USEPA to assess the health and sustainability of
ecosystems – forests cover 61% of the total land area in the MAIA region-. Ninety-five
percent of the region’s forests are classified as timberland. The vast majority (79%) of
timberlands is owned by nonindustrial private landowners, while the forest industry owns
approximately 7%. Hardwood forests dominate the MAIA region. For the region as a
whole, oak /hickory is the predominant forest type. Other dominant forest types in the
region are northern hardwoods, loblolly/shortleaf pine, and oak/pine.
4.4.1 Quantifying Forestry Benefits
Due to the different life cycles involved, the assessment of forest damage is substantially
more difficult than that of agricultural crop damage. On one hand trees live for a long
time, which makes the study of pollution impacts much more difficult. On the other hand,
-
The MAIA study region includes Delaware, Maryland, Pennsylvania, Virginia, and West Virginia, and
parts of New Jersey, New York, and North Carolina.
174
unlike agricultural soils, which are effectively managed through annual cycles, forest
soils are much less disturbed which leads to an accumulation of acidification impacts.
USEPA (1999) uses the PnET-II model to estimate the impacts of troposhperic ozone on
commercial timber growth. The PnET model simulates the cycles of carbon, water, and
nitrogen through forest ecosystems. Model inputs of monthly weather data and nitrogen
inputs are used to predict photosynthesis, evapotranspiration, and nitrogen cycling on a
monthly time-step for several forest types.
4.4.2 Valuation of Forestry Benefits
The valuation techniques of forestry benefits can be grouped into three categories:
1. direct market prices
2. indirect market prices
3. hypothetical values
Methods using direct market prices are based on actual prices, and consequently they do
not reflect some benefits (e.g. preservation of biodiversity) that the market participants
did not take into account in their decision-making. Methods utilizing indirect market
prices include hedonic property values, the travel costs, opportunity costs, surrogate
prices and replacement costs (Cavatassi, 2004). The opportunity cost method uses the
market price of the best alternative forgone to provide a lower bound on forestry benefits.
Surrogate prices methods use the market price of a close substitute as a proxy for the
benefits. A surrogate market approach is used by methods using hypothetical values. Two
methods that belong to this category are the contingent valuation method, and conjoint
analysis.
As described in the pervious sections, contingent valuation method uses surveys that ask
hypothetical questions to estimate economic values. Conjoint analysis estimates values
by asking people by asking people questions across a range of features or attributes of a
forestI. Forestry benefits may be grouped into three categories for valuation purposes: onsite private benefits, on-site public benefits, and global benefits. On-site private benefits
include timber productions, agricultural and other agroforestry products, and non-timber
forest products (e.g. mushrooms, medicinal plants, honey, fruits, nuts, etc.), and
recreation and tourism. On-site public benefits include watershed protection, agricultural
productivity enhancement, nutrient cycling, microclimate regulation, and aesthetic,
cultural, and spiritual values. Global benefits include carbon sequestration, and
biodiversity conservation.
Values derived for some of these benefits are not transferable, and therefore most
valuation studies restrict attention to on-site benefits such as timber production. These
benefits are usually estimated using market models. For example, USEPA (1999) used
I
Cavatassi (2004)
175
the USDA Forest Service Timber Assessment Market Model to estimate market changes
that result from reduced timber growth.
References
Abdalla, C.W., Roach, B.A., Epp, D.J. (1992). “Valuing environmental quality changes
using averting expenditure: an application to groundwater contamination”. Land
Economics 38 (2), 163–169.
Adams, R. M., Hamilton, S. A. and McCarl, B. A. (1986). “The benefits of pollution
control: the case of ozone and US agriculture”. American Journal of Agricultural
Economics 68, 886–889.
Ahn, Soeun; de Steiguer, Joseph E.; Palmquist, Raymond B.; Holmes, Thomas P. (2000).
“Economic Analysis of the Potential Impact of Climate Change on Recreational Trout
Fishing in the Southern Appalachian Mountains: An Application of a Nested Multinomial
Logit Model”, Climatic Change 45: 493-509.
Andreoni, James (1989). "Giving with Impure Altruism: Applications to Charity and
Ricardian Equivalence," Journal of Political Economy, Vol. 97 (6) pp. 1447-58.
University of Chicago Press
Armsworth, Paul R., Kendall, Bruce E., Davis, Frank W. (2004). “An Introduction to
biodiversity concepts for environmental economists” Resource and Energy Economics,
26, pp. 115-136
Ashmore, M. R. (1991). “Air Pollution and Agriculture”. Outlook on Agriculture 20,
139–144.
Bower, J.M., Stoll, J.R. (1988). “Use of dichotomous choice nonmarket methods to value
the whooping crane resource”. American Journal of Agricultural Economics 70, 327–
381.
Boyle, K.J. (1990). “Dichotomous choice, contingent valuation questions: functional
form is important”. Northeastern Journal of Agriculture and Resource Economics 19 (2),
125–131.
Boyle, K.J., Bishop, R.C. (1987). “Valuing wildlife in benefit cost analysis: a case study
involving endangered species”. Water Resources Research 23, 943–950.
Brookshire, D.S., Eubanks, L.S., Randall, A. (1983). “Estimating option prices and
existence values for wildlife resources”. Land Economics 59, 1–15.
176
Brown, William G. and Faisal M. Shalloof (1984). “Recommended Values for Columbia
River Salmon and Steelhead for Current Fishery Management Decisions.” Oregon State
University.
Burtraw, Dallas, Alan Krupnick, Erin Mansur, David Austin, and Deirdre Farrell (1997).
“The costs and benefits of reducing acid rain”, Discussion Paper 97-31- REV, Resources
for the Future, Washington, DC.
Cavatassi, R. (2004). “Valuation Methods for Environmental Benefits in Forestry and
Watershed Investment Projects”, The Food and Agriculture Organization of the United
Nations, ESA Working Paper No. 04-01.
Cosby, B.J., Hornberger, G.M., Galloway, J.N., Wright, R.F. (1985a). “Modelling the
effects of acid deposition: assessment of a lumped-parameter model of soil water and
streamwater chemistry”. Water Resour. Res., 21,51.
Cosby, B.J., Wright, R.F., Hornberger, G.M., Galloway, J.N. (1985b). “Modelling the
effects of acid deposition: estimation of long-term water quality responses in a small
forested catchment”. Water Resour. Res., 21, 1591.
Desvousges, W.H., Johnson, F.R., Dunford, R.W., Boyle, K.J., Hudson, S.P., Wilson,
K.N. (1993). “Measuring natural resource damages with contingent valuation: tests of
validity and reliability”, In: Hausman, J.A. (ed.), Contingent valuation: a critical
assessment, contributions to economic analysis, Chapter III, North-Holland, New York,
US.
Diamond, P.A., Hausman, J.A., Leonard, G.L., Denning, M.A. (1993). “Does contingent
valuation measure preferences? experimental evidence”, In: Hausman, J.A. (ed.),
Contingent valuation: a critical assessment, contributions to economic analysis, Chapter
II, North-Holland, New York, US.
Duffield, J.W., Patterson, D.A. (1992). “Field testing existence values: an instream flow
trust fund for Montana Rivers”, paper presented at the Allied Social Science Association
Annual Meeting, New Orleans, Louisiana.
Dutta, Nilima (1984). “The Value of Recreational Boating and Fishing in the Central
Basin of Ohio's Portion of Lake Erie.” Ohio State University, Sea Grant Program.
Feather, Peter M.; Hellerstein, Daniel, and Tomasi, Theodore (1995). “A Discrete-Count
Model of Recreational Demand”. Journal of Environmental Economics and Management.
29:214-227.
Freeman, A. Myrick, III. (1997). “On Valuing the Services and Functions of
Ecosystems.” EcosystemFunction and Human Activities: Reconciling Economics and
Ecology, eds. David R. Simpsonand Norman L. Christensen, Jr. New York: Chapman
and Hall.
177
Goulder, L.H., Kennedy, D. (1997). “Valuing ecosystem services: philosophical bases
and empirical methods”. In: Daily, G.C. (Ed.), Nature’s Services: Societal Dependence
on Natural Ecosystems. Island Press, Washington, DC, pp. 23–47.
Greene, Gretchen; Moss, Charles B., and Spreen, Thomas H. (1997). “Demand for
Recreational
Fishing in Tampa Bay, Florida: a Random Utility Approach”. Marine Resource
Economics. 12:293-305.
Haas, Mark A. and Stephen A. Weithman (1982). “Socioeconomic Value of the Trout
Fishery in Lake Teneycomo, Missouri.” Transactions of the American Fisheries Society
111, pp.223-230.
Hageman, R. (1985). “Valuing marine mammal populations: benefits valuations in
multispecies ecosystems”, In: Administrative Report, Southwest Fisheries, values cited in
Eagle, J., Betters, D., 1998. The Endangered Species Act and Economic Values: a
Comparison of Fines and Contingent Valuation Studies, Ecological Economics 26, 165–
171.
Halstead, J.M., Luloff, A.E., Stevens, T.H. (1992). “Protest bidders in contingent
valuation”. Northeastern Journal of Agricultural and Resource Economics 21 (2), 247–
361.
Herrick, C. N. and Kulp, J. L. (1987). “Interim Assessment: The Causes and Effects of
Acidic Deposition. Volume 1, Executive Summary”. Washington, DC: National Acid
Precipitation Assessment Program, Office for the Director of Research.
Hewitt, N. (1990). “How man-made ozone damages plants”. New Scientist, 25, pp. 25.
Hoehn, J.P., Loomis, J.B., (1993). “Substitution effects in the valuation of multiple
environmental programs”. Journal of Environmental Economics and Management 25 (1),
56–75.
Hoehn, J. P.; Tomasi, Theodore; Lupi, Frank, and Chen, Heng Z. (1996). “An Economic
Model for Valuing Recreational Angling Resources in Michigan”. Michigan State
University, Report to the Michigan Department of Environmental Quality.
Holland, M., Mills, G., Hayes, F., Buse, A., Emberson, L., Cambridge, H., Cinderby, H.,
Terry, A. and Ashmore, M. (2002). “Economic Assessment of Crop Yield Losses from
Ozone Exposure”. Part of the UNECE International Cooperative Programme on
Vegetation. Progress Report (April 2001 – March 2002). Contract EPG 1/3/170 . CEH
Project No: C01643.
Holmes, T.P. (1988). “The offsite impact of soil erosion on the water treatment industry”.
Land Economics 64 (4), 356–366.
178
Howitt, R. E. (1991). “Calibrated Models for Agricultural Production and Environmental
Analysis”. Working Paper 91-10. Davis, CA: University of California at Davis,
Department
of Agricultural Economics.
Howitt, R. E., Gossard, T. W. and Adams, R. M. (1984). “Effects of alternative ozone
concentrations and response data on economic assessments: the case of California crops”.
Journal of the Air Pollution Control Association 34, 1122–1127.
Huszar, P.C. (1989). “Economics of reducing off-site costs of wind erosion”. Land
Economics 65 (4), 333–340.
Jacobson, J. S. (1982). “Ozone and the growth and productivity of agricultural crops”. In
M.
H. Unsworth and D. P. Ormond (eds.), Effects of Gaseous Air Pollutants in Agriculture
and Horticulture, London: Butterworths.
Jakus, Paul M.; Dadakas, Dimitrios, and Fly, J. Mark (1998). “Fish Consumption
Advisories:
Incorporating Angler-Specific Knowledge, Habits, and Catch Rates in a Site Choice
Model”. American Journal of Agricultural Economics. 80(5):1019-1024.
Kealy, M.J., Turner, R.W. (1993). “A test of the equality of the closed-ended and the
open-ended contingent valuations”. American Journal of Agriculture Economics 75 (2),
311–331.
Kealy, Mary Jo and Richard C. Bishop (1986). “Theoretical and Empirical Specifications
Issues in Travel Cost Demand Studies.” American Journal of Agricultural Economics 68,
pp.660-667.
King, David A. and John G. Hof (1985). “Experimental Commodity Definition in
Recreation Travel Cost Methods.” Forest Science 31, pp. 519-529.
Laughland, A.S., Musser, W.N., Shortle, J.S., Musser, L.M. (1996). “Construct validity
of averting cost measures of environmental benefits”. Land Economics 72 (1), 100–112.
Lee, EH, WE Hogsett. (1996). “Methodology for calculating inputs for ozone secondary
standard benefits analysis: Part II”. Prepared for the US EPA, OAQPS.
Leung, S. K., Reed, W. and Geng, S. (1982). “Estimation of ozone damage to selected
crops grown in Southern California”. Journal of Air Pollution Control Association, 32,
pp. 160-164.
179
Loomis, J.B. (1989). “Test-retest reliability of the contingent valuation method: a
comparison of general population and visitor responses”. American Journal of
Agriculture Economics 71 (1), 76–84.
Loomis, J.B., Helfand, G. (1993). “A tale of two owls and lessons for the reauthorization
of the endangered species act”, Choices, 21–25.
Loomis, J.B., Larson, D.M. (1994). “Total economic value of increasing Gray Whale
populations: results from a contingent valuation survey of visitors and households”.
Marine Resource Economics 9, 275–286.
Loomis, John B. and Cindy F. Sorg (1986). “Economic Value of Idaho Sport Fisheries
With an Update on Valuation Techniques.” North American Journal of Fisheries
Management 6, pp.494-503.
Lupi, Frank and Hoehn, John P. (1998). “The Effect of Trip Lengths on Travel Cost
Parameters in Recreation Demand. Michigan State University”.
McConnell, K. E. and Strand, I. E. (1994). “The Economic Value of Mid and South
Atlantic
Sportfishing”. University of Maryland, Report to the USEPA and NOAA.
McClelland, G.H., Schulze, W.D, Lazo, J.K., Waldman, D.M., Doyle, J.K., Elliot, S.R.
and J.R. Irwin (1992). “Methods for measuring the non-use values: a contingent valuation
study of the groundwater cleanup”, draft report to USEPA, Center for Economic
Analysis, University of Colorado, Boulder, Colorado.
Miller, John R. and Michael J. Hay (1980). “Estimating Substate Values of Fishing and
Hunting.” Transactions of the North American Wilderness and Natural Resources
Conference 49(1980), pp.345-355.
Mitchell, R.C., Carson, R.T. (1984). “Willingness to pay for national freshwater
improvements, draft report to USEPA”, Resources for the Future, Washington, DC.
Montgomery, Mark and Needelman, Michael (1997). “The Welfare Effects of Toxic
Contamination in Freshwater Fish”. Land Economics. 73:211-273.
Morey, Edward R., Breffle, William S., and Greene, Pamela A. (2001). "Two Nested
Constant-Elasticity-of-Substitution Models of Recreational Participation and Site Choice:
An "Alternatives" Model and an "Expenditures" Model", American Journal of
Agricultural Economics. Vol 83, Issue 2, pp 414-427.
Murphy, J.J, Delucchi M.A., McCubbin D.R., Kim,H.J.(1999). “The Cost of Crop
Damage Caused by Ozone Air Pollution From Motor Vehicles”, Journal of
Environmental Management, vol. 55, pp. 273-289
180
Parsons, George R. and Hauber, A. Brett (1998). “Spatial Boundaries and Choice Set
Definition in a Random Utility Model of Recreation Demand”. Land Economics.
74(1):32-48.
Pendleton, Linwood H. and Mendelsohn, Robert (1998). “Estimating the Economic
Impact of
Climate Change on the Freshwater Sportsfisheries of the Northeastern U.S”. Land
Economics. 74(4):483-96.
Pimm, S.L., et al. (2001). “Can we defy Nature’s end?” Science 293, 2207–2208.
Reich, Peter B. and Robert G. Amundson (1985). “Ambient levels of ozone reduce net
photosynthesis in tree and crop species”. Science, vol. 230 1 November 1985 pp. 566-570
Ribaudo, M.O. (1989a). “Water quality benefits from the conservation reserve program,
agricultural economics report”, Economic Research Service, Washington.
Ribaudo, M.O. (1989b). “Targeting the conservation reserve program to maximize water
quality benefits”. Land Economics 65 (4), 320–332.
Richer, J. (1995). “Willingness to pay for desert protection”. Contemporary Economic
Policy XIII, 93–104.
Roughgarden, J. (1997). “Production functions from ecological populations: a survey
with emphasis on spatially explicit models”. In: Tilman, D. and P. Kareiva, (eds.) Spatial
Ecology: The Role of Space in rough Population Dynamics and Interspecific Interactions,
Princeton University Press, pp. 296-317.
Rowe, R. D. and Chestnut, L. G. (1985). “Economic assessment of the effects of air
pollution on agricultural crops in the San Joaquin Valley”. Journal of Air Pollution
Control Association, 35, 728-734.
Samples, K., Hollyer, J. (1989). “Contingent valuation of wildlife resources in the
presence of substitutes and complements”, In: Johnson, R., Johnson, G. (Eds.). Economic
Valuation of Natural Resources: Issues, Theory and Application, Boulder.
Schuhmann, Peter William (1998). “Deriving Species-Specific Benefits Measures for
Expected
Catch Improvements in a Random Utility Framework”, Marine Resource Economics,
Volume 13, No.1 Spring 1998, pp. 1-21.
Schluter, D., Ricklefs, R.E. (1993). “Species diversity: an introduction to the problem”.
In: Ricklefs, R.E., Schluter, D. (Eds.), Species Diversity in Ecological Communities.
University of Chicago Press, Chicago, pp. 1–10.
181
Segerson, K. (1987). “Economic impacts of ozone and acid rain: discussion”. American
Journal of Agricultural Economics, December, Volume 69, Number 5, pp. 970-971.
Segerson, K. (1991). “Air pollution and agriculture: a review and evaluation of policy
interactions”. In R. E. Just and N. Bockstael (eds.), Commodity and Resource Policies in
Agricultural Systems, Berlin: Springer-Verlag pp. 349-367, Chapter 18.
Silberman, J., Gerlowski, D.A., Williams, N.A. (1992). “Estimating existence value for
users and nonusers of New Jersey beaches”. Land Economics 68 (2), 225–236.
Smith, V.K., Desvousges, W.H. (1986). “Measuring Water Quality Benefits”. Kluwer
Nijhoff Publishing, Dordrecht.
Snyder, Lori D., Stavins, Robert, N., Wagner, Alexander F. (2003). “Private Options to
Use Public Goods: Exploiting Revealed Preferences to Estimate Environmental
Benefits”, Resource for the Future Discussion Paper 03-49.
Spash, C. L. (1997). “Assessing the Economic Benefits to Agriculture from Air Pollution
Control”, Journal of Economic Surveys, vol. 11, issue 1, pages 47-70
Stevens, T.H., DeCoteau, N.E., Willis, C.E., (1997). “Sensitivity of contingent valuation
to alternative payment schedules”. Land Economics 73 (1), 140–148.
Stoddard, J.L. et al. (1999). “Regional trends in aquatic recovery from acidification in
North America and Europe”. Nature, 401, 575-578.
Taylor, C.R. (1993). “AGSIM: An Econometric-Simulation Model of Regional Crop and
National Livestock Production in the United States”. In: C.R. Taylor, K.H. Reichelderfer,
and Train, Kenneth E. (1998). “Recreation Demand Models With Taste Differences Over
People”. Land Economics. 74(2):230-239.
Torell, L.A., Libbin, J.D., Miller, M.D. (1990). “The market value of water in the Ogalla
Aquifer”. Land Economics 66 (2), 163–175.
Turner R. K., Button, K., Nijkamp, P. (1999). Ecosystems and Nature: Economics,
Science and Policy, Edward Elgar Pub.
S.R. Johnson (Eds) Agricultural Sector Models for the United States: Descriptions and
Selected Policy Applications. Ames Iowa: Iowa State University Press.
U.S. EPA (1997). “Issue fact sheet on nutrients Report of Annual Meeting of National
Estuary Program Directors”, March 1997. National Estuary Program, Washington, D.C.
U.S. EPA, (1999). “The Benefits and Costs of the Clean Air Act: 1990 to 2010”. (EPA410-R-99-001), Office of Policy and Office of Air and Radiation, Washington, DC.
182
van Kooten, G.C., (1993). “Bioeconomic evaluation of government agricultural programs
on wetland conservation”. Land Economics 69 (1), 27–38.
Walker, D.J., Young, D.L. (1986). “The effect of technical progress erosion damage and
economic incentives for soil conservation”. Land Economics 62 (1), 83–93.
Walsh, Richard, Ray Ericson, Daniel Arosteguy, and Michael Hansen (1980). “An
Empirical Application of a Model for Estimating the Recreation Value of Instream
Flow.” Colorado Water Resources Research Institute, Colorado State University
Walsh, R.O., Loomis, J.B., Gillman, R.A. (1984). “Valuing option, existence, and
bequest demands for wilderness”. Land Economics 60, 14–29.
Wang, Deane; Bormann FH; and David Karnosky (1986). “Regional tree growth
reductions due to ambient ozone: evidence from field experiments”. Environ. Sci.
Technol., vol. 20 no 11, pp. 1122-1125.
Whitehead, J.C. (1993). “Total economic values for coastal and marine wildlife:
specification, validity and valuation issues”, Marine Resource Economics 8, 119–132.
Whittaker, R.H. (1972). “Evolution and measurement of species diversity”, Taxon 21,
213–251.
Ziemer, Rod, Wesley Musser, and Carter Hill (1980). “Recreation Demand Equations:
Functional Form and Consumer Surplus.” American Journal of Agricultural Economics
62, pp.136-141.
183
5. INDIRECT BENEFITS TO HUMANS THROUGH NONLIVING SYSTEMS
There are two types of indirect benefits through nonliving systems that were studied the
most: avoided materials damages, and improved visibility. The steps of estimating these
benefits are summarized in the sections that follow.
5.1 AVOIDED MATERIALS DAMAGES
Anthropogenic sulfur and nitrogen pollutants are believed to have caused vast damage to
buildings, structures, as well as the cultural heritage in the past century. Much of the
damage occurred in Europe and North America, but with growing car traffic, and very
high concentrations of sulfur dioxide in many cities of China, India, and Latin America
material damages due to air pollution continue to remain a significant problem.
Objects most susceptible to air pollution are the ones with long lives, particularly
buildings. Other objects, such as cars, may be damaged by air pollution, but these
damages tend to be less important, because they are usually replaced before the damage
could become significant.
Pollutants that contribute to degradation of buildings are particles (particularly soon)
causing soiling, and sulfur dioxide (SO2) contributing to corrosion and erosion caused by
acid rain.
Principal effects associated with air pollution are:
x
loss of mechanical strength
x
leakage,
x
failure of protective coatings,
x
loss of details in carvings,
x
pipe corrosion.
SO2 has a strong accelerating effect on the degradation of certain materials by
contributing to corrosion by acidic deposition. Atmospheric corrosion is influenced by
climatic patterns such as relative humidity, temperature and precipitation, and it tends to
be a local problem, because the damage often occurs near the source of emission. On the
other hand, indirect effects of SO2 emissions caused the acidification of soil and water
bodies, tend to be a regional problem due to the long-range transport of air pollutants.
Air pollution damages materials such as zinc, copper, stone, as well as organic materials.
In case of zinc and copper, the dissolution of protective corrosion products leads to
increased deterioration rates. Calcareous stones, such as limestone or marble, are very
susceptible to acid deposition by sulfur dioxide through transformation of the original
calcium carbonate to gypsum and calcium sulfate. Degradation of organic materials, such
184
as rubber tires and paints, are usually associated with ozone in conjunction with
temperature and solar radiation.
5.1.1 Quantifying materials damages
The ideal approach of quantifying materials damages is analogous to the approach used
to quantify health endpoints. One would start by estimating the change in pollutant
concentrations caused by a policy-induced reduction in emissions. The second step would
involve the use of dose-response or concentration-response (CR) functions that relate the
physical damage to ambient pollutant concentrations. And the last step involves attaching
monetary values to damages.
A valid CR function provides a mathematical relationship between properties of the
environment and some index of materials, such as loss of stock thickness. Some early
attempts aimed at deriving such CR functions focused on the relationship between
ambient pollutant concentrations and corrosion rates. As Lipfert (1996) points out, this
approach neglects the important variable of delivery of the reactant to the surface. A full
understanding of the process requires a separation of pollutant delivery process from the
subsequent chemical reactions. The appropriate technique to estimate a CR function is
that of a multiple regression, in which some index of corrosion is the independent
variable and the various environmental factors are the independent variables.
Perhaps the most difficult element of economic assessment of materials damages (or
benefits from reduced air pollution) is the problem of estimating stocks of buildings at
risk (Lipfert and Daum, 1992). One problem is that there is considerable heterogeneity in
the use of housing materials across country. While residential housing materials tend to
follow regional patterns, commercial and industrial buildings tend to be more uniform.
Another important problem since atmospheric corrosion has been present in many parts
of the country for a long time, people may have substituted away from the more sensitive
building materials toward less sensitive ones. The greatest difficulty lies in distinguishing
chemical characteristics of exposed surfaces within each building type and category.
Lipfert (1996) suggests that there is a need for a probabilistic, as opposed to
deterministic, approach to assessment. There are many relevant but disparate databases
on building stocks, but no effort has been made yet to synthesize that information
(Lipfert, 1996).
As an alternative to the above bottom-up approach, Rabl (1999) estimates damages to
buildings in France by working with aggregate data on observed frequencies of cleaning
and repair activities. The result is a “combined concentration-response function”. The
main variables in the CR function are income and the ambient concentration of
particulate matter.
Rabl (1999) considers two types of damage caused by air pollution:
x
Corrosion or erosion of coatings and contruction materials
185
x
Soiling
Corrosion and erosion are primarily due to acid deposition. A large number of studies
analyzed the effects of air pollution and corrosion and erosion. Dose-response functions
have been estimated for several building materials (e.g., Kucera, 1990; Haneef et al.,
1992; Butlin et al., 1992, Lipfert, 1987). There are relatively few studies on soiling due to
air pollution, and consequently few dose-response functions are available (Hamilton and
Mansfield, 1992).
5.1.2 Valuation of materials damages
Given a valid dose response function, using the bottom-up approach one would estimate
the repair cost due to air pollution as follows:
Total Repair Cost = Sum Surface Area * Repair Frequency * Repair Cost
Using a bottom-up approach the following are the steps in valuations:
x
Division into pollution strata
x
Materials inventory and inspection of physical damage
x
Damage functions
x
Estimated change in service life
x
Maintenance/ Replacement cost
x
Estimated economic damage
The main drawback of the bottom-approach is the need for very detailed data on building
inventories. As an alternative to the bottom-up approach, Rabl (1999) uses a linear
regression of renovation expenditures against income, PM13, and SO2. The best
regression model was the following:
R
E 0 E1 Income E 2 PM 13
where R and Income are measured in monetary units per person per year, while, PM13 is
the measure of particulate matter concentrations in Pg m 3 , and E0, E1 and E2 are
parameters to be estimated from the data. The above equation is what Rabl (1999) calls a
combined or aggregate concentration-response function. The change in repair costs in
response to a change in pollutant concentrations is then given by the following
expression:
186
wR
wPM 13
E 2 Income
Neither approach to valuation may be used to assess damages to the cultural heritage.
Cultural heritage encompasses both outdoor buildings and sculptures and treasured
objects kept indoors, stored in museums and archives. The most appropriate valuation
method for assessment is contingent valuation. These valuation studies tend to be case
specific and generally not transferable.
187
5.2 VISIBILITY BENEFITS
Reduced visibility due to anthropogenic air pollution affects some of the country’s most
scenic areas. US EPA estimates that in national parks in the eastern United States,
average visual range has decreased from 90 miles to 15-25 miles. In the West, visual
range has decreased from 140 miles to 35-90 miles. The main cause of visibility
impairment is haze. Under stagnant air mass conditions, aerosols can be trapped and
produce a visibility condition usually referred to as layered haze. Some light is absorbed
by particles while other light may be scattered away before it reaches the observer. The
introduction of particulate matter and certain gases into the atmosphere therefore reduces
visibility.
From a technical point of view, visibility is a complex and difficult concept to define.
Visibility includes psychophysical processes and concurrent value judgments of visual
impacts, as well as the physical interaction of light with particles in the atmosphere.
Therefore it is important to understand the psychological process involved in viewing a
scenic resource, and to be able to establish a link between the physical and psychological
processes.
5.2.1 Quantifying visibility benefits
Quantifying visibility requires developing links between visibility and particles that
scatter and absorb light. Visibility, in the most general sense, reduces to understanding
the effect that various types of aerosol and lighting conditions have on the appearance of
landscape features. Measuring visibility by a single index is, however, not possible
because visibility cannot be defined by a single parameter (Malm, 1999). Many visibility
indices have been proposed, however the most simple and direct way of communicating
reduced visibility is through a photograph. In fact, many contingent valuation (CV)
studies of visibility present the subjects with photographs of scenic areas with varying
levels of visibility. The reason photographs communicate visibility changes so well is
that the human eye works much like a camera. The human eye detects relative differences
in brightness rather than the overall brightness level, that is to say, the eye measures
contrast between adjacent objects.
Because the human eye function like a camera, a photograph captures visibility changes,
as humans perceive it. However, it is difficult to extract quantitative information from
photographs, and therefore direct measure of fundamental optical measures of the
atmosphere have been developed. The most common measures are atmospheric
extinction and scattering.
The scattering coefficient is a measure of the ability of particles to scatter photons out of
a beam of light, while the absorption coefficient is a measure of how many photons are
absorbed. Both coefficients are expressed as a number proportional to the amount of
photons scattered or absorbed per distance. The sum of scattering and absorption is
referred to as extinction or attenuation.
188
5.2.2 Valuation of visibility benefits
The most commonly used methods for visibility valuation are hedonic property values
and contingent valuation. Hedonic methods are based on revealed preference of
consumers, because they link nonmarket valuation to a traded commodity. Hedonic
property value studies estimate the marginal WTP function on the basis of an estimated
relationship between housing prices and housing attributes (including air quality). There
are several factors that affect the relationship between property values and air quality.
These include adverse health effects, reduced visibility or soiling due to air pollution.
Hedonic methods cannot be used to estimate separately. Disaggregation of overall
impacts requires making subjective judgments by the researcher. Nevertheless, the results
of hedonic property value studies confirm the hypothesis that air quality has a significant
impact on property prices. Kenneth and Greenstone (1998) estimate that the Clean Air
Act induced nationwide monetized benefits were $80 billion (in 1982-84 dollars) in the
1970’s, and $50 billion during the 1980s. Delucchi, Murphy and McCubbin (2002)
estimate monetized costs of total suspended particle pollution in 1990 at $52-$88 billion
in (1990 dollars). Some studies (e.g., Brookshire et al., 1979, 1982; Loehman et al., 1994;
McLelland et al., 1991) attempted to disaggregate property value impacts into health,
visibility, soiling, and other impacts. They find that visibility impacts are the second most
important, after health effects, representing 19-34% of total monetized benefits.
Burtraw et al. (1997) present the results of an integrated assessment of the benefits and
costs of the Title IV of the 1990 Clean Air Act Amendments initiated reductions in
emissions of sulfur dioxide and nitrogen oxides. They use the Tracking and Analysis
Framework (TAF) developed for the National Acid Precipitation Assessment Program
(NAPAP). Although uncertainties surround their estimates, the findings suggest that the
benefits of the program substantially outweigh its costs. Two types of visibility effects
are examined: recreational visibility at two national parks (Grand Canyon and
Shenandoah), and residential visibility in five metropolitan areas (Albany, NY, Atlantic
City, NJ, Charlottesville, VA, Knoxville, TN, and Washington, DC). The results,
summarized in Table 5.2.1 below, are most usefully considered on a per capita basis.
189
Table 5.2.1 Per Capita Benefits in 2010 for Affected Population
Effect
Benefits per Capita (1990$)
Morbidity
3.50
Mortality
59.29
Aquatic
0.62
Recreational Visibility
3.34
Residential Visibility
5.81
Costs
5.30
Source: Burtraw et al. (1997), Table 2, pp. 13
These visibility estimates illustrate their potential magnitude, but it should be noted that
they are based on relatively small number of studies available in the literature, and also
the geographical scope of the project is rather limited. Burtraw et al. (1997) explain the
relatively large magnitude of visibility benefits compared to other types of benefits,
namely aquatics, by claiming that willingness to pay depends on the availability of
substitutes, and visibility, along with health, has no close substitutes.
Smith and Osborne (1996) perform a meta-analysis of visibility valuation studies to test
whether CV estimates of WTP are responsive to the amount, or scope, of the
environmental amenity being offered. They consider an internal consistency test for CVbased WTP. Internal consistency tests assess the reliability and validity of CV surveys.
On way to evaluate the CV method is to compare willingness to pay WTP functions
estimated with CV surveys with the specific, observable properties that economic theory
implies WTP should follow. Smith and Osborne (1996) selected five of CV studies that
used comparable methods for the meta-analysis. These studies focused on air quality as a
key element. Furthermore, in each study air quality is presented in a way that permits
computation of the change in visible range. The five selected studies are summarized in
Table 5.2.2 below.
190
Table 5.2.2 Summary of CV studies for visibility at national parks analyzed by Smith and Osborne
(1996)
Mean change
Location
Type of survey
Authors
Mean and interin visibility
quartile range of
WTP
(per month in 1990
$)
Rowe et al. (1980)
MacFarland et al
$9.27
($6.83, $10.82)
$2.75
($1.69, $3.73)
0.50
1.18
Schulze et al.
$8.50
($4.42, $11.67)
0.62
Balson et al.
$0.46
($0.007, $0.97)
Grand Canyon and
Mesa Verde
National Parks
In-person interviews
administered to
households in
Albuquerque, Los
Angeles, Denver,
and Chicago
Grand Canyon,
Yosemite, and
Shenandoah
National Parks
Mail with telephone
households in
Arizona, Virginia,
California, New
York, and Missouri
conducted in St.
Louis and San
Diego Counties
Grand Canyon
National Park
0.955
administered to to
households in area
administered to
visitors to the area
Grand Canyon,
Mesa Verde, and
Zion National Parks
0.79
Chestnut and Rowe
$4.35
($3.15, $5.48)
Navaho Recreation
Area
Source: Smith and Osborne (1996), pp. 291, Table 1
The findings of Smith and Osborne (1996) support a positive, statistically significant and
robust relationship between the WTP estimates and the percentage improvement in
visible range. These results suggest that it may be possible to transfer results from a metaanalysis of past CV studies. The crucial issue in benefit transfer is to find a common
metric to measure the environmental amenity.
191
References:
Balson, W. E., Carson, R. T., Conaway, M. B., Fischhoff, B., Hanemann, W. M., Hulse,
A., Kopp, R. J., Martin, K. M., Mitchell, R. C., Molenar, J., Presser, S., and Ruud, P. A.
(1990). ‘‘Development and Design of a Contingent Valuation Survey for Measuring the
Public’s Value for Visibility Improvements at the Grand Canyon National Park,’’ draft
report, Decision Focus, Inc., Los Altos, CA.
Butlin, RN et al. (1992). “Preliminary results from the analysis of stone tablets from the
National Materials Exposure Program (NMEP)”, Atmospheric Environment, vol. 26B,
number 2, pp. 189-198
Chestnut, L. G. and Rowe, R. D. (1990). ‘‘Preservation Value for Visibility Protection at
the National Parks,’’ Draft final report to U.S. Environmental Protection Agency,
RCGrHagler Bailley
Delucchi, Mark, James Murphy,Donald McCubbin (2002). “The Health and Visibility
Cost of Air Pollution: A Comparison of Estimation Methods” Journal of Environmental
Management 64, 139–152
Haneef, SJ, Johnson, JB, Dickinson, C, Thompson, GE, Wood, GC (1992). “Effect of
Dry Deposition of NOx and SO2 gaseous pollutants on the degradation of calcareous
building stones”, Atmospheric Environment, vol. 26A, number 16, pp. 2963-2974
Kenneth Y. Chay, Michael Greenstone (1998). “Does Air Quality Matter? Evidence from
the Housing Market” NBER Working Paper No. w6826
Kucera, V. (1990). “External Building Materials – Quantities and Degradation”, The
National Swedish Institute for Building Research, Gaole, Sweden, Research Report
TN:19
Lipfert F. W. (1987). “Estimating the effects of air pollution on buildings and structures:
the U.S. experience since 1985 and some lessons for the future”, Brookhaven National
Laboratory
MacFarland, K., W. Malm, and Molenar, J. (1983). “An examination of methodologies
for assessing the value of visibility”, in ‘‘Managing Air Quality and Scenic Resources at
National Parks and Wilderness Areas’’ R. Rowe and L. Chestnut, Eds. , Westview Press,
Boulder, CO.
Rabl, A. (1999). “Air Pollution and Buildings: An Estimation of Damage Costs in
France”, Environmental Impact Assessment Review, 1999, vol. 19 (4), pp. 361-385
Rowe, R. D., d’Arge, R., and Brookshire, D. (1980). “An experiment on the economic
value of visibility”, Journal of Environmental Economics and Management, Vol. 7, pp. 119
192
Tasdemiroglu, F. (1991). “Costs of Material Damage Due to Air Pollution in the United
States”, Renewable Energy, 1991, 1, pp. 639-48
Schulze, W. D., Brookshire, D., Walther, E. and Kelly, K. (1983). ‘‘The Benefits of
Preserving Visibility in the National Parklands of the Southwest, VIII,’’ draft report
prepared for the U.S. Environmental Protection Agency, Washington, D.C.
193
6. SAMPLE CALCULATIONS
The following examples illustrate the steps and calculations involved in the estimation of
non-market benefits of the RPS. The examples below focus on mortality and morbidity
benefits only.
Step1: Projecting Concentrations
In this step the change in pollutant concentrations around each power plant is estimated.
Instead of using a more sophisticated dispersion model, we used the Gaussian Plume
Dispersion Model (GPDM) to project pollutant concentrations. This model predicts an
average concentration under steady state conditions. The shape of the plume undergoing
dispersion is a function of the wind speed, vertical temperature profile and atmospheric
stability. GPDM is widely used to predict concentrations in the atmosphere. There are,
however, significant simplifications with this model. The main assumptions of GPDM
are:
1. Only steady-state concentrations are estimated.
2. Wind blows in x-direction and is constant in both speed and direction
3. Transport with the mean wind is much greater than turbulent transport in the xdirection.
4. The source emission rate (the rate at which the pollutant is emitted per unit of time) is
constant.
5. Diffusion coefficients are constant in both time and space.
6. The source emits chemicals of concern (COC) at a point in space x=y=0 and z=H,
where H is the effective stack height.
7. The COC’s are inert (non-decaying and non-reactive).
8. There is no barrier to plume migration.
9. Mass is conserved across the plume cross section.
10. Mass within a plume follows a Gaussian (normal) distribution in both the crosswind
(y) and vertical (z) directions.
The GPDM is derived from the advection-diffusion equation. The general equation to
calculate the steady state concentration of an air contaminant in the ambient air resulting
from a point source is given by:
C(y, x, z)
§
·
§
·½
¨ y 2 ¸° §¨ (z H) 2 ·¸
¨ (z H) 2 ¸°
Q
exp¨
¸ exp
¸ exp¨
¸¾
2 ¸®° ¨¨ 2ı 2
2ʌʌu ı
¨
¸
¨ 2ı 2
¸°
2ı
y z
z
z
¹
©
¹¿
© y ¹¯ ©
where
C(x,y,z) = contaminant concentration at the specified coordinate
x = downwind distance
194
y = crosswind distance
z = vertical distance above ground
Q = contaminant emission rate
Vy = lateral dispersion coefficient function
Vz = vertical dispersion coefficient function
u = wind velocity in downwind direction
H = effective stack height
In the above equation Vy, the lateral dispersion coefficient function, and Vz, the vertical
dispersion coefficient functions depend on the downwind distance and the atmospheric
stability class. The value of these coefficients in meters can be obtained from the
equations utilized by the Industrial Source Complex (ISC) Dispersion Model developed
by USEPA (1995):
ı
465.11628 x tan(TH)
y
where
TH
ı
z
0.01745 >c dln(x)@
ax b
A simplified version of the above formula to estimate steady state pollutant
concentrations is given by
C(y, x, z)

§
·½
° 1 ¨ y 2 (z H) 2 ¸°
Q
exp® ¨
¸
2 ¸¾°
2ʌʌu ı
2 ¨ ı2
ı
°
y z
z ¹¿
© y
¯
This formula assumes that the pollutant is not reflected from the ground, and therefore it
yields lower estimates in general than when one assumes reflection.
195
EXAMPLE 1: Reduction in excess mortality due to SO2
Power plant assumptions
We have in mind a coal-burning power plant with the following characteristics:
Capacity = 800MW
Capacity Factor = 0.7M
SO2 emission factor per output: 10 lbs/MWh
Contaminant emissions rate (g/s): 705
Physical stack height (m): 125
Effective stack height (m): 215.96
Consistent with the followings assumptions:
Stack velocity (m/s): 15
Stack exit diameter (m): 1.5
Stack gas temperature (K): 450
Ambient gas temperature (K): 300
We assume that the RPS results in a 10% reduction in generation by this power plant.
Assumptions about population and health status
Total population: 8.4 million
Total non-accidental deaths: 70,766
Non-accidental mortality per person: 0.00841
Population is assumed to be uniformly distributed in the affected area.
Population density per square kilometer: 402
Meteorological and other assumptions
Wind velocity in downward direction (m/s): 1
Incoming solar radiation during the day: moderate
PASQUILL-GIFFORD category: B
Cloud condition at night: mostly overcast
PASQUILL-GIFFORD category: B
Step 2: Quantifying the change in SO2 mortality
First, using GPDM we calculated SO2 concentrations for 100-meter grids for 25 km
downwind and 9 km crosswind distances. Next, the estimated concentrations were
M
Capacity factor is the ratio of the electrical energy produced by a generating unit for the period of time
considered to the electrical energy that could have been produced at continuous full power operation during
the same period.
196
averaged at the 1-km grid level. These steps were followed for both the baseline scenario
and the policy scenario (10% reduction in generation). The difference in SO2
concentration between the two scenarios at the 1-km grid level is estimated.
Reduction in SO2 Concentrations
Concentration
(ug/m3)
50.00
40.00-50.00
30.00-40.00
20.00-30.00
10.00-20.00
0.00-10.00
40.00
30.00
20.00
10.00
25
21
17
13
9
5
S13
1
0.00
S1
Crosswind
distance (km)
Downwind distance (km)
Next concentration response functions are used to calculate the change in non-accidental
mortality.
CR-function:
ǻmortality
ȕǻSO
§
2 1) ·¸ pop
¨ y1 (e
¸
¨
¹
©
y1
=
non-accidental deaths per person
E
=
SO2 coefficient
=
=
population sample
'SO
pop
We used the SO2 mortality CR function estimated by Touloumi et al. (1996). The 95%
confidence interval (CI) for the reduction in non-accidental mortality (statistical deaths)
is: (27.5, 72.2)1 with an estimated value 30.89.
1
This CI refers only to the uncertainty associated with the CR-function estimation.
197
Step 3: Monetizing SO2 mortality benefits
Using the median estimate of Viscusi and Aldy (2003) for the Value of Statistical Life
(VSL) of $7 million, the estimated benefit is $216.2 million with a 95% confidence
interval of ($192.5m, 505.4m).
EXAMPLE 2: Reduction in excess chronic obstructive pulmonary disease (COPD)
hospital admissions due to PM10
This example illustrates the estimation of morbidity benefits associated with improved air
quality. The health benefit of interest is avoided hospital admissions for chronic
obstructive pulmonary disease (COPD) due to particulate matter 10 Pm and less in
diameter (PM10). COPD is a group of diseases categorized by ICD-9-CM codes 490-496.
(ICD=International Classification of Diseases)
490 - Bronchitis, not specified as acute or chronic
491 - Chronic bronchitis
492 - Emphysema
493 - Asthma
494 - Bronchiectasis
495 - Extrinsic allergic alveolitis
496 - Chronic airway obstruction, not elsewhere classified
The same power plant, population and meteorological assumptions hold as I the previous
example. Morbidity benefits were estimated in the following steps.
Step 1: Determining concentration-response (CR) relationships for COPD admissions.
CR-studies usually assume the following log-linear functional form:
ȕPM
§
10 1) ·¸ pop
ǻCOPD hospital admissions - ¨ y (e
¨ 1
¸
©
¹
= baseline COPD admission rate, defined as COPD hospital admissions per person
y1
E estimatedPM10 coefficient
198
'PM10 = change in PM10 concentrations (Pg/m3)
pop
= exposed population per km2
Three studies have been identified that estimated the concentration-response relationship
between 'PM10 concentrations and hospital admissions for COPD: Chen at al. (2004) in
Vancouver, Canada, Zanobetti et al. (2000) in Chicago, Cook county, IL, and
Moolgavkar (2000) in Los Angeles county, CA. The parameter estimates are summarized
in the table below.
Study
Parameter estimate 95% Confidence Interval
EL
EH
E
Chen et al. (2004)
0.0152463
0.0061760 0.0243167
Zanobetti et al.
(2000)
0.0076035
0.0015873 0.0136196
0.0016073
0.0010603 0.0021542
Moolgavkar
0.0007968
0.0002110 0.0016826
(2000)
0.0009877
0.0004969 0.0014785
Study
Population
Ages 65+
Studied Health
Effect
ICD-9
490-492,494,496
Ages 65+
Ages 0-19
Ages 20-64
Ages 65+
490-492, 494-496
490-496
490-496
490-496
Step 2: Determining Baseline Exposure
We use the Gaussian Plume Dispersion Model to predict PM10 concentration changes.
Under our meteorological assumptions most of the deposition occurs within 25 km in
downwind direction and 10 km in crosswind direction. We assume that population is
uniformly distributed in the affected area. We use a population density estimate derived
from 2003 population estimate figures by the U.S. Census Bureau: 1164 people per
square mile, or 450 people per square kilometer. Consequently, the total population that
is potentially exposed to PM pollutant from the power plant is approximately 250,000.
Based on U.S. Census Bureau population estimates, we assume the following population
density estimate for the various age groups:
Age Group
Share of Total Population
Population Density per Square
Kilometer
0-19
27.2%
122.1
20-64
59.7%
268.3
65+
13.2%
59.4
Step 3: Determining the Number of Baseline Cases for Each Quantifiable Health
Effect Number exposed x Baseline exposure x Dose-response relationship.
199
Because the health effects studied in the three studies are not identical, it was necessary
to estimate and make assumptions about the baseline hospital admission rate for each
group of health effects (ICD codes). The Healthcare Cost & Utilization Project (HCUP)
by
the
Agency
for
Healthcare
Research
and
Quality's
(AHRQ)
(http://www.ahrq.gov/hcupnet/), estimated the number of hospital discharges for the
various COPD conditions by various characteristics, such as age, sex, income, etc., at the
national and regional (but not state) level. Because our CR-functions are estimated for
various age groups, it was necessary to estimate the hospital admission rate for each age
group. Hospital admissions were estimated for the following age groups from the 2002
national data:
Age Group
Share of Total
ICD 490-492,
ICD 490-492,
ICD 490-496
Population
494,496
494-496
0-17
25.3%
0.00006174
0.00006174
0.0020353
18-64
62.3%
0.00112073
0.00112073
0.0021734
65+
12.4%
0.0116307
0.0116307
0.0136903
Step 4: Determining Exposure for each Policy Scenario, Determining the Number of
Cases for Each Quantifiable Effect with the Regulation, Determining the Number of
Cases Avoided as a Result of Each Regulatory Option
The avoided COPD hospital admissions for the various age groups and CR functions and
dispersion models are summarized in the tables below.
Estimated
Reduction in
COPD Hospital
Admissions
Dispersion Model: Gaussian Plume Dispersion without Reflection
95% Confidence CR-Function
Population
Location
Interval
Chen et al.
Ages 65+
(2004)
Vancouver, Canada
Zanobetti et al. Ages 65+ (Medicare
2.98
0.60
5.55
(2000)
admissions)
Chicago, Cook County, IL
Moolgavkar
Ages 65+
0.44
0.22
0.66
(2000)
Los Angeles County
Moolgavkar
0.25
0.07
0.54
Ages 20-64
(2000)
Los Angeles County
Moolgavkar
Ages 0-19
0.22
0.14
0.29
(2000)
Los Angeles County
Dispersion Model: Gaussian Plume Dispersion with Reflection
CR-Function
Population
Location
Estimated Reduction 95% Confidence
Interval
in COPD Hospital
Admissions
6.28
15.81
200
2.40
10.66
5.68
28.88 Chen et al. (2004)
Ages 65+
Vancouver, Canada
7.12
1.38
13.81
Zanobetti et al.
(2000)
Ages 65+
1.00
0.50
1.51
Moolgavkar (2000)
Ages 65+
0.58
0.15
1.24
Moolgavkar (2000) Ages 20-64
0.50
0.33
0.68
Moolgavkar (2000) Ages 0-19
Chicago, Cook County, IL
Los Angeles County
Los Angeles County
Los Angeles County
Step 5: Monetizing the health benefits
Benefits were monetized using the cost-of-illness approach. Using HCUP estimates of
mean cost of hospital admissions, weighted cost was estimated for each age group and
each groups of ICD codes. The following value were used to monetize benefits:
Age Group
ICD-9 Codes
Ages 65+
Ages 65+
Ages 65+
Ages 19-64
Ages 0-18
490-496
490-492, 494-496
490-492, 494, 496
490-496
490-496
Average Cost
($2002)
15,537.05
13,908.82
13,886.41
12,421.90
7,511.36
201
Monetized benefits for the two dispersion models are summarized in the tables below.
Dispersion without
reflection
Chen et al. (2004)
Moolgavkar (2000)
Population
65+
65+
Medicare patients
All ages
Population
Dispersion with
relfection
Chen et al. (2004)
Moolgavkar (2000)
65+
65+
Medicare patients
All ages
Avoided COPD
Monetary
Admissions Value ($2002)
(annual)
87,186.3
6.3
3.0
41,455.3
0.9
11,583.5
Avoided COPD
Monetary
Admissions Value ($2002)
(annual)
219,559.5
15.8
95% Confidence interval
33,319.7
8,340.3
147,969.3
77,177.5
5,319.3
19,087.0
95% Confidence interval
78,847.9
401,042.0
7.1
98,996.9
19,198.5
192,047.0
2.1
26,578.3
12,173.5
43,946.7
References:
Chen Y, Yang Q, Krewski D, Shi Y, Burnett RT, McGrail K. Influence of relatively low
level of particulate ar pollution on hospitalization for COPD in elderly people. Inhal
Toxicol. 2004 Jan;16(1):21-5.
ISC3 (Industrial Source Complex Model), User’s Guide, Volume 2, page 1-16, Table 1-1,
http://www.epa.gov/scram001/tt22.htm
Moolgavkar SH. Air pollution and hospital admissions for chronic obstructive pulmonary
disease in three metropolitan areas in the United States. Inhal Toxicol. 2000;12 Suppl
4:75-90.
Zanobetti A, Schwartz J, Gold D. Are there sensitive subgroups for the effects of airborne
particles? Environ Health Perspect. 2000 Sep;108(9):841-5
Moolgavkar (2000)
202
Center for Energy, Economic and Environmental Policy
Advisory Board
As of December 1, 2004
Dennis Bone, President, Verizon New Jersey
Donald DiFrancesco, Former Governor, New Jersey
Laurence Downes, Chairman & Chief Executive Officer, New Jersey Resources
Zulima Farber, Esquire, Lowenstein Sandler PC
James Florio, Chairman & Chief Executive Officer, XSPAND
Jeffrey Genzer, Esquire,General Council, Association of State Energy Officials,
Duncan, Weinberg, Genzer & Pembroke
Edward Graham, President, South Jersey Gas Company
Ashok Gupta, Director of Air and Energy Programs, Natural Resource Defense Council
Ralph Izzo, President & Chief Operating Officer, PSE&G
Alfred Koeppe, President, Newark Alliance
Steven Morgan, President, Jersey Central Power and Light
Gary Stockbridge, Vice President of Customer Relationship Management, Conectiv Power Delivery
Carol Werner, Executive Director, Environmental and Energy Study Institute
203
3VUHFST5IF4UBUF6OJWFSTJUZPG/FX+FSTFZ
-JWJOHTUPO"WFOVF
/FX#SVOTXJDL/FX+FSTFZ